mirror of
https://github.com/rustfs/rustfs.git
synced 2026-08-16 18:08:21 +00:00
Compare commits
1 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
 overtrue
|
195f19217f |
@@ -252,16 +252,10 @@ test-group = 'ecstore-serial-flaky'
|
||||
# cluster, so it keeps the lane's parallel-safe / no-external-dependency
|
||||
# properties. The RustFS warm backend has no loopback guard (that guard is
|
||||
# replication-only), so it needs no opt-in env for its 127.0.0.1 tier target.
|
||||
#
|
||||
# Disk compression (backlog#1848): the `compression` module joins the smoke
|
||||
# lane so the multipart disk-compression roundtrips (restored after
|
||||
# rustfs/rustfs#5169 disabled them) have PR-lane signal, not just merge-gate.
|
||||
# Single-node servers on random ports with isolated temp dirs — meets the
|
||||
# admission criteria unchanged.
|
||||
[profile.e2e-smoke]
|
||||
default-filter = """
|
||||
package(e2e_test) & (
|
||||
test(/^(delete_marker_migration_semantics|version_id_regression|list_objects_v2_pagination|list_object_versions_regression|list_objects_duplicates|list_buckets_double_slash|list_buckets_auth|list_buckets_iam_filter|leading_slash_key|special_chars|create_bucket_region|delete_objects_versioning|head_object_consistency|head_object_range|copy_object_metadata|copy_object_tagging|copy_source_invalid_date|content_encoding|compression|multipart_storage_class|storage_class_capability|ssec_copy|anonymous_access|bucket_policy_check|presigned_negative|negative_sigv4|admin_auth|notification_webhook|tls_hot_reload|console_smoke|admin_iam_crud|admin_pools|sts_query_compat)_test::|^fake_s3_target::/)
|
||||
test(/^(delete_marker_migration_semantics|version_id_regression|list_objects_v2_pagination|list_object_versions_regression|list_objects_duplicates|list_buckets_double_slash|list_buckets_auth|list_buckets_iam_filter|leading_slash_key|special_chars|create_bucket_region|delete_objects_versioning|head_object_consistency|head_object_range|copy_object_metadata|copy_object_tagging|copy_source_invalid_date|content_encoding|multipart_storage_class|storage_class_capability|ssec_copy|anonymous_access|bucket_policy_check|presigned_negative|negative_sigv4|admin_auth|notification_webhook|tls_hot_reload|console_smoke|admin_iam_crud|admin_pools|sts_query_compat)_test::|^fake_s3_target::/)
|
||||
| test(/^replication_extension_test::(test_replication_check_succeeds_with_remote_target|test_replication_check_rejects_target_without_object_lock|test_set_remote_target_rejects_unversioned_source_bucket|test_replication_check_rejects_unversioned_source_bucket|test_replication_check_rejects_missing_replication_config|test_replication_check_rejects_invalid_bucket|test_set_remote_target_rejects_same_bucket_on_same_deployment|test_set_remote_target_rejects_unversioned_target_bucket|test_set_remote_target_update_requires_arn|test_set_remote_target_update_rejects_missing_target|test_set_remote_target_rejects_invalid_target_url|test_set_remote_target_rejects_self_signed_https_target_without_skip_tls_verify|test_set_remote_target_rejects_private_ca_https_target_without_ca_cert_pem|test_list_remote_targets_rejects_empty_bucket|test_list_remote_targets_rejects_invalid_bucket|test_remove_remote_target_rejects_missing_target|test_remove_remote_target_rejects_missing_arn|test_remove_remote_target_rejects_invalid_bucket|test_remove_remote_target_rejects_target_used_by_replication|test_delete_bucket_replication_removes_remote_target)$/)
|
||||
| test(/^reliant::lifecycle::/)
|
||||
| test(/^reliant::tiering::/)
|
||||
|
||||
@@ -85,7 +85,7 @@ runs:
|
||||
repo-token: ${{ github.token }}
|
||||
|
||||
- name: Install flatc
|
||||
uses: Nugine/setup-flatc@698800de72a96bfb22cf60431dc21a2ff9a7e07b # v1
|
||||
uses: Nugine/setup-flatc@e7855e994773ce90094a3f1626d4afc9080c23ae # v1
|
||||
with:
|
||||
version: "25.12.19"
|
||||
|
||||
|
||||
@@ -182,12 +182,7 @@ jobs:
|
||||
echo '```'
|
||||
} >> "$GITHUB_STEP_SUMMARY"
|
||||
|
||||
# Readers: test-and-lint-rio-v2 (per-PR), build-rustfs-debug-binary-rio-v2
|
||||
# (weekly schedule / manual dispatch only — dormant rio-v2 variant, see
|
||||
# rustfs/backlog#1835 and docs/architecture/minio-file-format-compat.md).
|
||||
# The second build below stays despite the reduced cadence: it warms the
|
||||
# rio-v2,e2e-test-hooks feature resolution the scheduled build restores,
|
||||
# which keeps that lane inside its 30-minute timeout.
|
||||
# Readers: test-and-lint-rio-v2, build-rustfs-debug-binary-rio-v2.
|
||||
warm-ci-feat-rio:
|
||||
name: Warm ci-feat-rio
|
||||
runs-on: sm-standard-4
|
||||
|
||||
@@ -533,12 +533,7 @@ jobs:
|
||||
|
||||
build-rustfs-debug-binary-rio-v2:
|
||||
name: Build RustFS Debug Binary (rio-v2)
|
||||
# Dormant rio-v2 variant (rustfs/backlog#1835): the feature ships in no
|
||||
# default build, so this full-suite lane runs only on the weekly schedule
|
||||
# and manual dispatch. Per-PR cfg-seam coverage stays with
|
||||
# test-and-lint-rio-v2. Lifecycle and the promote-or-delete condition:
|
||||
# docs/architecture/minio-file-format-compat.md ("rio-v2 variant lifecycle").
|
||||
if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'
|
||||
if: github.event_name != 'pull_request' || github.event.action != 'closed'
|
||||
needs: [ quick-checks ]
|
||||
runs-on: sm-standard-4
|
||||
timeout-minutes: 30
|
||||
@@ -829,9 +824,6 @@ jobs:
|
||||
|
||||
e2e-tests-rio-v2:
|
||||
name: End-to-End Tests (rio-v2)
|
||||
# Inherits the schedule/dispatch-only gate through needs: on every other
|
||||
# event build-rustfs-debug-binary-rio-v2 is skipped, so this job skips
|
||||
# with it (see the dormant-variant comment on that job).
|
||||
needs: [ build-rustfs-debug-binary-rio-v2 ]
|
||||
runs-on: sm-standard-2
|
||||
timeout-minutes: 30
|
||||
|
||||
@@ -94,7 +94,6 @@ jobs:
|
||||
short_sha: ${{ steps.check.outputs.short_sha }}
|
||||
is_prerelease: ${{ steps.check.outputs.is_prerelease }}
|
||||
create_latest: ${{ steps.check.outputs.create_latest }}
|
||||
source_ref: ${{ steps.check.outputs.source_ref }}
|
||||
steps:
|
||||
- name: Checkout repository
|
||||
uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7
|
||||
@@ -119,7 +118,6 @@ jobs:
|
||||
short_sha=""
|
||||
is_prerelease=false
|
||||
create_latest=false
|
||||
source_ref="$GITHUB_SHA"
|
||||
|
||||
if [[ "${{ github.event_name }}" == "workflow_run" ]]; then
|
||||
# Triggered by build workflow completion
|
||||
@@ -139,7 +137,6 @@ jobs:
|
||||
# Extract version info from commit message or use commit SHA
|
||||
# Use Git to generate consistent short SHA (ensures uniqueness like build.yml)
|
||||
short_sha=$(git rev-parse --short "$HEAD_SHA")
|
||||
source_ref="$HEAD_SHA"
|
||||
|
||||
# Determine build type based on triggering workflow event and ref
|
||||
triggering_event="$TRIGGERING_EVENT"
|
||||
@@ -264,23 +261,6 @@ jobs:
|
||||
echo "⚠️ Only release versions (latest, v1.0.0, 1.0.0) and prereleases (v1.0.0-alpha1, 1.0.0-beta2) are supported"
|
||||
;;
|
||||
esac
|
||||
|
||||
if [[ "$should_build" == true && "$input_version" != "latest" ]]; then
|
||||
tag_ref="refs/tags/$input_version"
|
||||
if ! git ls-remote --exit-code origin "$tag_ref" >/dev/null 2>&1; then
|
||||
if [[ "$input_version" == v* ]]; then
|
||||
tag_ref="refs/tags/${input_version#v}"
|
||||
else
|
||||
tag_ref="refs/tags/v$input_version"
|
||||
fi
|
||||
fi
|
||||
|
||||
if ! git ls-remote --exit-code origin "$tag_ref" >/dev/null 2>&1; then
|
||||
echo "❌ Release tag not found for Docker build: $input_version"
|
||||
exit 1
|
||||
fi
|
||||
source_ref="$tag_ref"
|
||||
fi
|
||||
fi
|
||||
|
||||
{
|
||||
@@ -291,7 +271,6 @@ jobs:
|
||||
echo "short_sha=$short_sha"
|
||||
echo "is_prerelease=$is_prerelease"
|
||||
echo "create_latest=$create_latest"
|
||||
echo "source_ref=$source_ref"
|
||||
} >> "$GITHUB_OUTPUT"
|
||||
|
||||
echo "🐳 Docker Build Summary:"
|
||||
@@ -302,7 +281,6 @@ jobs:
|
||||
echo " - Short SHA: $short_sha"
|
||||
echo " - Is prerelease: $is_prerelease"
|
||||
echo " - Create latest: $create_latest"
|
||||
echo " - Source ref: $source_ref"
|
||||
|
||||
# Build multi-arch Docker images
|
||||
# Strategy: Build images using pre-built binaries from dl.rustfs.com
|
||||
@@ -330,7 +308,6 @@ jobs:
|
||||
uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7
|
||||
with:
|
||||
persist-credentials: false
|
||||
ref: ${{ needs.build-check.outputs.source_ref }}
|
||||
|
||||
- name: Login to Docker Hub
|
||||
uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # v3
|
||||
@@ -420,8 +397,7 @@ jobs:
|
||||
LABELS="org.opencontainers.image.title=RustFS"
|
||||
LABELS="$LABELS,org.opencontainers.image.description=RustFS distributed object storage system"
|
||||
LABELS="$LABELS,org.opencontainers.image.version=$VERSION"
|
||||
SOURCE_REVISION="$(git rev-parse HEAD)"
|
||||
LABELS="$LABELS,org.opencontainers.image.revision=$SOURCE_REVISION"
|
||||
LABELS="$LABELS,org.opencontainers.image.revision=${{ github.sha }}"
|
||||
LABELS="$LABELS,org.opencontainers.image.source=${{ github.server_url }}/${{ github.repository }}"
|
||||
LABELS="$LABELS,org.opencontainers.image.created=$(date -u +'%Y-%m-%dT%H:%M:%SZ')"
|
||||
LABELS="$LABELS,org.opencontainers.image.build-type=$BUILD_TYPE"
|
||||
|
||||
+4
-7
@@ -101,10 +101,7 @@ refactors.
|
||||
|
||||
The `rustfs` binary crate composes these libraries into the running server.
|
||||
`ecstore` remains the storage engine at the architectural center; its internal
|
||||
module split is tracked under `docs/architecture/`. `rio-v2` is the
|
||||
feature-gated MinIO on-disk format compatibility I/O layer; it ships in no
|
||||
default build (lifecycle:
|
||||
[docs/architecture/minio-file-format-compat.md](docs/architecture/minio-file-format-compat.md)).
|
||||
module split is tracked under `docs/architecture/`.
|
||||
|
||||
## Architecture Invariants
|
||||
|
||||
@@ -134,9 +131,9 @@ default build (lifecycle:
|
||||
why it stays local).
|
||||
- ✅ RESOLVED: `BackpressureConfig` and `DataUsageInfo` each have exactly one
|
||||
definition (`crates/io-core/src/backpressure.rs`,
|
||||
`crates/data-usage/src/data_usage.rs`). The zero-consumer
|
||||
`BackpressureSettings` copy that lingered in io-metrics was removed
|
||||
(rustfs/backlog#1833).
|
||||
`crates/data-usage/src/data_usage.rs`). A zero-consumer
|
||||
`BackpressureSettings` copy lingers in `crates/io-metrics/src/config.rs`;
|
||||
its removal is tracked in rustfs/backlog#1833.
|
||||
|
||||
4. **ecstore does not know about HTTP or S3 protocol details.** It operates on
|
||||
storage-level abstractions (objects, buckets, disks, pools).
|
||||
|
||||
Generated
+78
-90
@@ -278,7 +278,6 @@ checksum = "312c1ea69e5fe9966e0029fb95aca8790100b85aff4f0d3b00a9337c74069a9c"
|
||||
dependencies = [
|
||||
"bigdecimal",
|
||||
"bon",
|
||||
"crc32fast",
|
||||
"digest 0.11.3",
|
||||
"log",
|
||||
"miniz_oxide 0.9.1",
|
||||
@@ -290,11 +289,9 @@ dependencies = [
|
||||
"serde",
|
||||
"serde_bytes",
|
||||
"serde_json",
|
||||
"snap",
|
||||
"strum",
|
||||
"thiserror 2.0.20",
|
||||
"uuid",
|
||||
"zstd",
|
||||
]
|
||||
|
||||
[[package]]
|
||||
@@ -1165,9 +1162,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "aws-smithy-eventstream"
|
||||
version = "0.61.2"
|
||||
version = "0.61.1"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "6de526c7b567420a31bc283657a7921b45c4cafe0827fdf2490713dcc770c28f"
|
||||
checksum = "5a9381123ab62d20c13082b151f30f962a3b112b727345394536dfa39a482944"
|
||||
dependencies = [
|
||||
"aws-smithy-types",
|
||||
"bytes",
|
||||
@@ -1198,9 +1195,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "aws-smithy-http-client"
|
||||
version = "1.3.0"
|
||||
version = "1.2.0"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "3c1c8a04cb31ba74d0115af5a890bb8c0d48fba64b52812fa13929a6ef0cc83c"
|
||||
checksum = "635d23afda0a6ab48d666c4d447c4873e8d1e83518a2be2093122397e50b838e"
|
||||
dependencies = [
|
||||
"aws-smithy-async",
|
||||
"aws-smithy-protocol-test",
|
||||
@@ -1280,9 +1277,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "aws-smithy-runtime"
|
||||
version = "1.13.1"
|
||||
version = "1.12.1"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "483b858ff67522011c4786310c5cd8fd88d0be7ea3d5f1a48328446300c4269e"
|
||||
checksum = "07505b34e8f4b3591a4fa69e9792b52289b95488dbbc68c3c0075b7bedb245e1"
|
||||
dependencies = [
|
||||
"aws-smithy-async",
|
||||
"aws-smithy-http",
|
||||
@@ -1346,9 +1343,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "aws-smithy-types"
|
||||
version = "1.6.2"
|
||||
version = "1.6.1"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "fce83ce9abbb198d25bc7131e468d0f9fe1257125e58c39f3f9fc9f5098c9647"
|
||||
checksum = "d6dc683efb34b9e755675b37fedbe0103141e5b6df7bdc9eb6967756a8c167d8"
|
||||
dependencies = [
|
||||
"base64-simd",
|
||||
"bytes",
|
||||
@@ -3764,7 +3761,7 @@ checksum = "d0881ea181b1df73ff77ffaaf9c7544ecc11e82fba9b5f27b262a3c73a332555"
|
||||
|
||||
[[package]]
|
||||
name = "e2e_test"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"anyhow",
|
||||
"astral-tokio-tar",
|
||||
@@ -5136,9 +5133,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "http-body-util"
|
||||
version = "0.1.5"
|
||||
version = "0.1.4"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "23169fe34a5fbcdd3f3862e78fb9b6fccd5f02a6dc6f732547005d45631ce71c"
|
||||
checksum = "e9f41fd6a08e4d4ec69df65976da761afd5ad5e58a9d4acb46bd1c953a9e3ff2"
|
||||
dependencies = [
|
||||
"bytes",
|
||||
"futures-core",
|
||||
@@ -5960,7 +5957,7 @@ checksum = "b6d2cec3eae94f9f509c767b45932f1ada8350c4bdb85af2fcab4a3c14807981"
|
||||
[[package]]
|
||||
name = "libmimalloc-sys"
|
||||
version = "0.1.49"
|
||||
source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11#6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11"
|
||||
source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=ce6338661179c8be22e516b00af7483f151485a7#ce6338661179c8be22e516b00af7483f151485a7"
|
||||
dependencies = [
|
||||
"cc",
|
||||
"cty",
|
||||
@@ -6262,9 +6259,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "metrique"
|
||||
version = "0.1.30"
|
||||
version = "0.1.29"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "dedbf06ffeef4c37990c73636fbd993aa34fb1948afd736e6114f239220993db"
|
||||
checksum = "d2e394c63e2d1a30aeb3b9392ecf3439d8475d2df810a8f4f6e66d6866754017"
|
||||
dependencies = [
|
||||
"itoa",
|
||||
"jiff",
|
||||
@@ -6292,9 +6289,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "metrique-macro"
|
||||
version = "0.1.21"
|
||||
version = "0.1.20"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "f4fb1f30185f53f7f6e4c9e46745c1a1350af8e77fda5a88aded44b0637a82e0"
|
||||
checksum = "786df1fd0abebd0db685f7e9a353c78756d4b370fb98a52376c2015fa55f141f"
|
||||
dependencies = [
|
||||
"Inflector",
|
||||
"darling 0.23.0",
|
||||
@@ -6321,9 +6318,9 @@ checksum = "2faca4e4480069ff02b1763b3b79f5cec7e8628e24d9dc5b6073f53d2577a4d9"
|
||||
|
||||
[[package]]
|
||||
name = "metrique-writer"
|
||||
version = "0.1.26"
|
||||
version = "0.1.25"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "20bd17c1a3ca2719e31f19ce77a853948dc2102f35976b92276c42a64fdc5f3f"
|
||||
checksum = "82cdde44d241dab7fc8b7a32e0eb5dae6cd28f8de80b59f9a1e9f2f0b05e485e"
|
||||
dependencies = [
|
||||
"ahash",
|
||||
"crossbeam-queue",
|
||||
@@ -6342,9 +6339,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "metrique-writer-core"
|
||||
version = "0.1.20"
|
||||
version = "0.1.19"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "f1a55b6aae1d85c557c729564c4e2b32a26dc65ba2d90d9647ca01f2bd4854c4"
|
||||
checksum = "e57379b7ee2272efaeaaa6de062503563e57333b24aadc7f2255b3d602899e8b"
|
||||
dependencies = [
|
||||
"derive-where",
|
||||
"itertools 0.14.0",
|
||||
@@ -6369,7 +6366,7 @@ dependencies = [
|
||||
[[package]]
|
||||
name = "mimalloc"
|
||||
version = "0.1.52"
|
||||
source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11#6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11"
|
||||
source = "git+https://github.com/xonatius/mimalloc_rust.git?rev=ce6338661179c8be22e516b00af7483f151485a7#ce6338661179c8be22e516b00af7483f151485a7"
|
||||
dependencies = [
|
||||
"libmimalloc-sys",
|
||||
]
|
||||
@@ -6885,7 +6882,7 @@ version = "5.0.0"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "51e219e79014df21a225b1860a479e2dcd7cbd9130f4defd4bd0e191ea31d67d"
|
||||
dependencies = [
|
||||
"base64 0.21.7",
|
||||
"base64 0.22.1",
|
||||
"chrono",
|
||||
"getrandom 0.2.17",
|
||||
"http 1.5.0",
|
||||
@@ -8046,7 +8043,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "be769465445e8c1474e9c5dac2018218498557af32d9ed057325ec9a41ae81bf"
|
||||
dependencies = [
|
||||
"heck 0.5.0",
|
||||
"itertools 0.10.5",
|
||||
"itertools 0.14.0",
|
||||
"log",
|
||||
"multimap",
|
||||
"once_cell",
|
||||
@@ -8066,7 +8063,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "03da047801ff44bb6a4d407d4860c05fd70bb81714e6b2f3812603d5b145b042"
|
||||
dependencies = [
|
||||
"heck 0.5.0",
|
||||
"itertools 0.10.5",
|
||||
"itertools 0.14.0",
|
||||
"log",
|
||||
"multimap",
|
||||
"petgraph 0.8.3",
|
||||
@@ -8087,7 +8084,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "8a56d757972c98b346a9b766e3f02746cde6dd1cd1d1d563472929fdd74bec4d"
|
||||
dependencies = [
|
||||
"anyhow",
|
||||
"itertools 0.10.5",
|
||||
"itertools 0.14.0",
|
||||
"proc-macro2",
|
||||
"quote",
|
||||
"syn 2.0.119",
|
||||
@@ -8100,7 +8097,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "b570b25f7617e43d59005d0990ccb79e950a423952cea19671b7a876da390adf"
|
||||
dependencies = [
|
||||
"anyhow",
|
||||
"itertools 0.10.5",
|
||||
"itertools 0.14.0",
|
||||
"proc-macro2",
|
||||
"quote",
|
||||
"syn 2.0.119",
|
||||
@@ -9093,7 +9090,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"aes-gcm",
|
||||
"anyhow",
|
||||
@@ -9203,8 +9200,8 @@ dependencies = [
|
||||
"serial_test",
|
||||
"sha2 0.11.0",
|
||||
"shadow-rs",
|
||||
"snap",
|
||||
"socket2",
|
||||
"starshard",
|
||||
"subtle",
|
||||
"sysinfo",
|
||||
"temp-env",
|
||||
@@ -9231,7 +9228,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-audit"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"const-str",
|
||||
@@ -9254,7 +9251,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-checksums"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"base64-simd",
|
||||
"bytes",
|
||||
@@ -9270,7 +9267,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-common"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"chrono",
|
||||
"hotpath",
|
||||
@@ -9288,7 +9285,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-concurrency"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"insta",
|
||||
@@ -9301,7 +9298,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-config"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"const-str",
|
||||
"hotpath",
|
||||
@@ -9311,7 +9308,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-credentials"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"base64-simd",
|
||||
"hmac 0.13.0",
|
||||
@@ -9325,7 +9322,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-crypto"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"aes-gcm",
|
||||
"argon2",
|
||||
@@ -9346,7 +9343,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-data-usage"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"rmp-serde",
|
||||
@@ -9356,7 +9353,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-ecstore"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"async-channel",
|
||||
@@ -9495,7 +9492,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-extension-schema"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"serde",
|
||||
@@ -9505,7 +9502,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-filemeta"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"byteorder",
|
||||
@@ -9532,7 +9529,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-heal"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"base64 0.23.1",
|
||||
@@ -9563,7 +9560,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-iam"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"async-trait",
|
||||
@@ -9604,7 +9601,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-io-core"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"bytes",
|
||||
"hotpath",
|
||||
@@ -9617,14 +9614,16 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-io-metrics"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"criterion",
|
||||
"hotpath",
|
||||
"metrics",
|
||||
"metrics-util",
|
||||
"num_cpus",
|
||||
"rustfs-common",
|
||||
"rustfs-s3-ops",
|
||||
"rustfs-utils",
|
||||
"sysinfo",
|
||||
"thiserror 2.0.20",
|
||||
"tokio",
|
||||
@@ -9681,7 +9680,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-keystone"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"bytes",
|
||||
"futures",
|
||||
@@ -9708,7 +9707,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-kms"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"aes-gcm",
|
||||
"anyhow",
|
||||
@@ -9738,7 +9737,6 @@ dependencies = [
|
||||
"rustfs-utils",
|
||||
"rustify",
|
||||
"serde",
|
||||
"serde_ignored",
|
||||
"serde_json",
|
||||
"sha2 0.11.0",
|
||||
"subtle",
|
||||
@@ -9757,7 +9755,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-lifecycle"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"hotpath",
|
||||
@@ -9780,7 +9778,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-lock"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"compact_str",
|
||||
@@ -9803,7 +9801,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-log-analyzer"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"chrono",
|
||||
"flate2",
|
||||
@@ -9822,7 +9820,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-madmin"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"humantime",
|
||||
@@ -9837,7 +9835,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-notify"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"async-trait",
|
||||
@@ -9872,7 +9870,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-object-capacity"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"criterion",
|
||||
"futures",
|
||||
@@ -9892,7 +9890,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-object-data-cache"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"bytes",
|
||||
"criterion",
|
||||
@@ -9909,7 +9907,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-obs"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"chrono",
|
||||
"crossbeam-channel",
|
||||
@@ -9964,7 +9962,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-policy"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"base64-simd",
|
||||
@@ -9995,7 +9993,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-protocols"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"astral-tokio-tar",
|
||||
"async-compression",
|
||||
@@ -10057,7 +10055,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-protos"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"flatbuffers",
|
||||
"hotpath",
|
||||
@@ -10081,7 +10079,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-replication"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"byteorder",
|
||||
"bytes",
|
||||
@@ -10099,7 +10097,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-rio"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"aes-gcm",
|
||||
"arc-swap",
|
||||
@@ -10137,7 +10135,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-rio-v2"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"aes-gcm",
|
||||
"bytes",
|
||||
@@ -10160,7 +10158,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-s3-ops"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"rustfs-s3-types",
|
||||
@@ -10168,7 +10166,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-s3-types"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"serde",
|
||||
@@ -10177,7 +10175,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-s3select-api"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"bytes",
|
||||
@@ -10207,7 +10205,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-s3select-query"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-recursion",
|
||||
"async-trait",
|
||||
@@ -10226,7 +10224,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-scanner"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"bytes",
|
||||
@@ -10266,7 +10264,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-security-governance"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"thiserror 2.0.20",
|
||||
@@ -10274,7 +10272,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-signer"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"base64-simd",
|
||||
"bytes",
|
||||
@@ -10292,7 +10290,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-storage-api"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"hotpath",
|
||||
@@ -10307,7 +10305,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-targets"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"async-nats",
|
||||
@@ -10361,7 +10359,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-test-utils"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"hotpath",
|
||||
"rustfs-data-usage",
|
||||
@@ -10377,7 +10375,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-tls-runtime"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"arc-swap",
|
||||
"hotpath",
|
||||
@@ -10398,7 +10396,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-trusted-proxies"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"async-trait",
|
||||
"axum",
|
||||
@@ -10435,7 +10433,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-utils"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"base64-simd",
|
||||
"blake2",
|
||||
@@ -10477,7 +10475,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustfs-zip"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
dependencies = [
|
||||
"astral-tokio-tar",
|
||||
"async-compression",
|
||||
@@ -10563,9 +10561,9 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "rustls-connector"
|
||||
version = "0.23.8"
|
||||
version = "0.23.7"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "1babecfcc65b139b812e74bcc7f9ec7b4e00db659fd42d99567b7e77f0c714c6"
|
||||
checksum = "09a5abe04eec18f8b9fbe87885bcaee6426de80bbc579958c0bc064b728ee617"
|
||||
dependencies = [
|
||||
"futures-io",
|
||||
"futures-rustls",
|
||||
@@ -10941,16 +10939,6 @@ dependencies = [
|
||||
"syn 3.0.3",
|
||||
]
|
||||
|
||||
[[package]]
|
||||
name = "serde_ignored"
|
||||
version = "0.1.14"
|
||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "115dffd5f3853e06e746965a20dcbae6ee747ae30b543d91b0e089668bb07798"
|
||||
dependencies = [
|
||||
"serde",
|
||||
"serde_core",
|
||||
]
|
||||
|
||||
[[package]]
|
||||
name = "serde_json"
|
||||
version = "1.0.151"
|
||||
@@ -11804,7 +11792,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||
checksum = "32497e9a4c7b38532efcdebeef879707aa9f794296a4f0244f6f69e9bc8574bd"
|
||||
dependencies = [
|
||||
"fastrand",
|
||||
"getrandom 0.3.4",
|
||||
"getrandom 0.4.3",
|
||||
"once_cell",
|
||||
"rustix",
|
||||
"windows-sys 0.61.2",
|
||||
|
||||
+54
-55
@@ -41,7 +41,7 @@ members = [
|
||||
"crates/protocols", # Protocol implementations (FTPS, SFTP, etc.)
|
||||
"crates/protos", # Protocol buffer definitions
|
||||
"crates/rio", # Rust I/O utilities and abstractions
|
||||
"crates/rio-v2", # MinIO on-disk format compatibility I/O layer (feature-gated, ships in no default build)
|
||||
"crates/rio-v2", # Next-generation Rust I/O compatibility layer
|
||||
"crates/replication", # Replication contracts and wire formats
|
||||
"crates/concurrency", # Concurrency management for RustFS - timeout, locking, backpressure, and I/O scheduling
|
||||
"crates/s3-types", # S3 event type definitions
|
||||
@@ -69,7 +69,7 @@ edition = "2024"
|
||||
license = "Apache-2.0"
|
||||
repository = "https://github.com/rustfs/rustfs"
|
||||
rust-version = "1.97.1"
|
||||
version = "1.0.0-rc.2"
|
||||
version = "1.0.0-rc.1"
|
||||
homepage = "https://rustfs.com"
|
||||
description = "RustFS is a high-performance distributed object storage software built using Rust, one of the most popular languages worldwide. "
|
||||
keywords = ["RustFS", "Minio", "object-storage", "filesystem", "s3"]
|
||||
@@ -86,52 +86,52 @@ redundant_clone = "warn"
|
||||
|
||||
[workspace.dependencies]
|
||||
# RustFS Internal Crates
|
||||
rustfs = { path = "./rustfs", version = "1.0.0-rc.2" }
|
||||
rustfs-heal = { path = "crates/heal", version = "1.0.0-rc.2" }
|
||||
rustfs-audit = { path = "crates/audit", version = "1.0.0-rc.2" }
|
||||
rustfs-checksums = { path = "crates/checksums", version = "1.0.0-rc.2" }
|
||||
rustfs-common = { path = "crates/common", version = "1.0.0-rc.2" }
|
||||
rustfs-data-usage = { path = "crates/data-usage", version = "1.0.0-rc.2" }
|
||||
rustfs-config = { path = "./crates/config", version = "1.0.0-rc.2" }
|
||||
rustfs-concurrency = { path = "./crates/concurrency", version = "1.0.0-rc.2" }
|
||||
rustfs-credentials = { path = "crates/credentials", version = "1.0.0-rc.2" }
|
||||
rustfs-crypto = { path = "crates/crypto", version = "1.0.0-rc.2" }
|
||||
rustfs-ecstore = { path = "crates/ecstore", version = "1.0.0-rc.2" }
|
||||
rustfs-filemeta = { path = "crates/filemeta", version = "1.0.0-rc.2" }
|
||||
rustfs-iam = { path = "crates/iam", version = "1.0.0-rc.2" }
|
||||
rustfs-keystone = { path = "crates/keystone", version = "1.0.0-rc.2" }
|
||||
rustfs-lifecycle = { path = "crates/lifecycle", version = "1.0.0-rc.2" }
|
||||
rustfs-kms = { path = "crates/kms", version = "1.0.0-rc.2" }
|
||||
rustfs-lock = { path = "crates/lock", version = "1.0.0-rc.2" }
|
||||
rustfs-madmin = { path = "crates/madmin", version = "1.0.0-rc.2" }
|
||||
rustfs-notify = { path = "crates/notify", version = "1.0.0-rc.2" }
|
||||
rustfs-io-metrics = { path = "crates/io-metrics", version = "1.0.0-rc.2" }
|
||||
rustfs-io-core = { path = "crates/io-core", version = "1.0.0-rc.2" }
|
||||
rustfs-object-capacity = { path = "crates/object-capacity", version = "1.0.0-rc.2" }
|
||||
rustfs-object-data-cache = { path = "crates/object-data-cache", version = "1.0.0-rc.2", default-features = false }
|
||||
rustfs-log-analyzer = { path = "crates/log-analyzer", version = "1.0.0-rc.2" }
|
||||
rustfs-obs = { path = "crates/obs", version = "1.0.0-rc.2" }
|
||||
rustfs-policy = { path = "crates/policy", version = "1.0.0-rc.2" }
|
||||
rustfs-protos = { path = "crates/protos", version = "1.0.0-rc.2" }
|
||||
rustfs-protocols = { path = "crates/protocols", version = "1.0.0-rc.2" }
|
||||
rustfs-replication = { path = "crates/replication", version = "1.0.0-rc.2" }
|
||||
rustfs-rio = { path = "crates/rio", version = "1.0.0-rc.2" }
|
||||
rustfs-rio-v2 = { path = "crates/rio-v2", version = "1.0.0-rc.2" }
|
||||
rustfs-s3-types = { path = "crates/s3-types", version = "1.0.0-rc.2" }
|
||||
rustfs-s3-ops = { path = "crates/s3-ops", version = "1.0.0-rc.2" }
|
||||
rustfs-s3select-api = { path = "crates/s3select-api", version = "1.0.0-rc.2" }
|
||||
rustfs-s3select-query = { path = "crates/s3select-query", version = "1.0.0-rc.2" }
|
||||
rustfs-scanner = { path = "crates/scanner", version = "1.0.0-rc.2" }
|
||||
rustfs-security-governance = { path = "crates/security-governance", version = "1.0.0-rc.2" }
|
||||
rustfs-extension-schema = { path = "crates/extension-schema", version = "1.0.0-rc.2" }
|
||||
rustfs-signer = { path = "crates/signer", version = "1.0.0-rc.2" }
|
||||
rustfs-storage-api = { path = "crates/storage-api", version = "1.0.0-rc.2" }
|
||||
rustfs-trusted-proxies = { path = "crates/trusted-proxies", version = "1.0.0-rc.2" }
|
||||
rustfs-targets = { path = "crates/targets", version = "1.0.0-rc.2" }
|
||||
rustfs-test-utils = { path = "crates/test-utils", version = "1.0.0-rc.2" }
|
||||
rustfs-tls-runtime = { path = "crates/tls-runtime", version = "1.0.0-rc.2" }
|
||||
rustfs-utils = { path = "crates/utils", version = "1.0.0-rc.2" }
|
||||
rustfs-zip = { path = "./crates/zip", version = "1.0.0-rc.2" }
|
||||
rustfs = { path = "./rustfs", version = "1.0.0-rc.1" }
|
||||
rustfs-heal = { path = "crates/heal", version = "1.0.0-rc.1" }
|
||||
rustfs-audit = { path = "crates/audit", version = "1.0.0-rc.1" }
|
||||
rustfs-checksums = { path = "crates/checksums", version = "1.0.0-rc.1" }
|
||||
rustfs-common = { path = "crates/common", version = "1.0.0-rc.1" }
|
||||
rustfs-data-usage = { path = "crates/data-usage", version = "1.0.0-rc.1" }
|
||||
rustfs-config = { path = "./crates/config", version = "1.0.0-rc.1" }
|
||||
rustfs-concurrency = { path = "./crates/concurrency", version = "1.0.0-rc.1" }
|
||||
rustfs-credentials = { path = "crates/credentials", version = "1.0.0-rc.1" }
|
||||
rustfs-crypto = { path = "crates/crypto", version = "1.0.0-rc.1" }
|
||||
rustfs-ecstore = { path = "crates/ecstore", version = "1.0.0-rc.1" }
|
||||
rustfs-filemeta = { path = "crates/filemeta", version = "1.0.0-rc.1" }
|
||||
rustfs-iam = { path = "crates/iam", version = "1.0.0-rc.1" }
|
||||
rustfs-keystone = { path = "crates/keystone", version = "1.0.0-rc.1" }
|
||||
rustfs-lifecycle = { path = "crates/lifecycle", version = "1.0.0-rc.1" }
|
||||
rustfs-kms = { path = "crates/kms", version = "1.0.0-rc.1" }
|
||||
rustfs-lock = { path = "crates/lock", version = "1.0.0-rc.1" }
|
||||
rustfs-madmin = { path = "crates/madmin", version = "1.0.0-rc.1" }
|
||||
rustfs-notify = { path = "crates/notify", version = "1.0.0-rc.1" }
|
||||
rustfs-io-metrics = { path = "crates/io-metrics", version = "1.0.0-rc.1" }
|
||||
rustfs-io-core = { path = "crates/io-core", version = "1.0.0-rc.1" }
|
||||
rustfs-object-capacity = { path = "crates/object-capacity", version = "1.0.0-rc.1" }
|
||||
rustfs-object-data-cache = { path = "crates/object-data-cache", version = "1.0.0-rc.1", default-features = false }
|
||||
rustfs-log-analyzer = { path = "crates/log-analyzer", version = "1.0.0-rc.1" }
|
||||
rustfs-obs = { path = "crates/obs", version = "1.0.0-rc.1" }
|
||||
rustfs-policy = { path = "crates/policy", version = "1.0.0-rc.1" }
|
||||
rustfs-protos = { path = "crates/protos", version = "1.0.0-rc.1" }
|
||||
rustfs-protocols = { path = "crates/protocols", version = "1.0.0-rc.1" }
|
||||
rustfs-replication = { path = "crates/replication", version = "1.0.0-rc.1" }
|
||||
rustfs-rio = { path = "crates/rio", version = "1.0.0-rc.1" }
|
||||
rustfs-rio-v2 = { path = "crates/rio-v2", version = "1.0.0-rc.1" }
|
||||
rustfs-s3-types = { path = "crates/s3-types", version = "1.0.0-rc.1" }
|
||||
rustfs-s3-ops = { path = "crates/s3-ops", version = "1.0.0-rc.1" }
|
||||
rustfs-s3select-api = { path = "crates/s3select-api", version = "1.0.0-rc.1" }
|
||||
rustfs-s3select-query = { path = "crates/s3select-query", version = "1.0.0-rc.1" }
|
||||
rustfs-scanner = { path = "crates/scanner", version = "1.0.0-rc.1" }
|
||||
rustfs-security-governance = { path = "crates/security-governance", version = "1.0.0-rc.1" }
|
||||
rustfs-extension-schema = { path = "crates/extension-schema", version = "1.0.0-rc.1" }
|
||||
rustfs-signer = { path = "crates/signer", version = "1.0.0-rc.1" }
|
||||
rustfs-storage-api = { path = "crates/storage-api", version = "1.0.0-rc.1" }
|
||||
rustfs-trusted-proxies = { path = "crates/trusted-proxies", version = "1.0.0-rc.1" }
|
||||
rustfs-targets = { path = "crates/targets", version = "1.0.0-rc.1" }
|
||||
rustfs-test-utils = { path = "crates/test-utils", version = "1.0.0-rc.1" }
|
||||
rustfs-tls-runtime = { path = "crates/tls-runtime", version = "1.0.0-rc.1" }
|
||||
rustfs-utils = { path = "crates/utils", version = "1.0.0-rc.1" }
|
||||
rustfs-zip = { path = "./crates/zip", version = "1.0.0-rc.1" }
|
||||
|
||||
# Async Runtime and Networking
|
||||
async-channel = "2.5.0"
|
||||
@@ -154,7 +154,7 @@ hyper-rustls = { default-features = false, version = "0.27.9" }
|
||||
hyper-util = { version = "0.1.20" }
|
||||
http = "1.5.0"
|
||||
http-body = "1.1.0"
|
||||
http-body-util = "0.1.5"
|
||||
http-body-util = "0.1.4"
|
||||
minlz = "1.2.3"
|
||||
reqwest = "0.13.4"
|
||||
rustfs-kafka-async = { version = "1.2.0" }
|
||||
@@ -171,7 +171,7 @@ tower = { version = "0.5.3" }
|
||||
tower-http = { version = "0.7.0" }
|
||||
|
||||
# Serialization and Data Formats
|
||||
apache-avro = { version = "0.22.0", features = ["snappy", "zstandard"] }
|
||||
apache-avro = "0.22.0"
|
||||
bytes = { version = "1.12.1" }
|
||||
bytesize = "2.7.0"
|
||||
byteorder = "1.5.0"
|
||||
@@ -182,7 +182,6 @@ quick-xml = "0.41.0"
|
||||
rmp = { version = "0.8.15" }
|
||||
rmp-serde = { version = "1.3.1" }
|
||||
serde = { version = "1.0.229" }
|
||||
serde_ignored = { version = "0.1" }
|
||||
serde_json = { version = "1.0.151" }
|
||||
serde_urlencoded = "0.7.1"
|
||||
|
||||
@@ -231,9 +230,9 @@ aws-credential-types = { version = "1.3.0" }
|
||||
aws-sdk-kms = { default-features = false, version = "1.114.0" }
|
||||
aws-sdk-s3 = { default-features = false, version = "1.141.0" }
|
||||
aws-sdk-sts = { default-features = false, version = "1.110.0" }
|
||||
aws-smithy-http-client = { default-features = false, version = "1.3.0" }
|
||||
aws-smithy-http-client = { default-features = false, version = "1.2.0" }
|
||||
aws-smithy-runtime-api = { version = "1.14.0" }
|
||||
aws-smithy-types = { version = "1.6.2" }
|
||||
aws-smithy-types = { version = "1.6.1" }
|
||||
base64 = "0.23.1"
|
||||
base64-simd = "0.8.0"
|
||||
brotli = "8.0.4"
|
||||
@@ -348,8 +347,8 @@ russh-sftp = "2.4.0"
|
||||
dav-server = "0.11.0"
|
||||
|
||||
# Performance Analysis and Memory Profiling
|
||||
mimalloc = { version = "0.1.52", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11" }
|
||||
libmimalloc-sys = { version = "0.1.49", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "6d4c41bb10c6d9da1d1b6f07b38c4cc051667f11", features = ["extended"] }
|
||||
mimalloc = { version = "0.1.52", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "ce6338661179c8be22e516b00af7483f151485a7" }
|
||||
libmimalloc-sys = { version = "0.1.49", git = "https://github.com/xonatius/mimalloc_rust.git", rev = "ce6338661179c8be22e516b00af7483f151485a7", features = ["extended"] }
|
||||
hotpath = { version = "0.23.2", default-features = false }
|
||||
# Snapshot testing for output format regression detection
|
||||
insta = { version = "1.48" }
|
||||
|
||||
@@ -116,7 +116,7 @@ chown -R 10001:10001 data logs
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:latest
|
||||
|
||||
# Using specific version
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:1.0.0-rc.2
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:1.0.0-rc.1
|
||||
```
|
||||
|
||||
If you use [podman](https://github.com/containers/podman) instead of docker, you can install the RustFS with the below command
|
||||
|
||||
+1
-1
@@ -113,7 +113,7 @@ chown -R 10001:10001 data logs
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:latest
|
||||
|
||||
# 使用指定版本运行
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:1.0.0-rc.2
|
||||
docker run -d -p 9000:9000 -p 9001:9001 -v $(pwd)/data:/data -v $(pwd)/logs:/logs rustfs/rustfs:1.0.0-rc.1
|
||||
```
|
||||
|
||||
如果您通过绑定挂载启用 TLS 证书目录，也请用同样方式准备该目录：
|
||||
|
||||
@@ -11,6 +11,7 @@
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
#![allow(dead_code)]
|
||||
|
||||
use base64_simd::STANDARD;
|
||||
|
||||
|
||||
@@ -21,13 +21,6 @@ use crate::{
|
||||
Xxhash3, Xxhash64, Xxhash128,
|
||||
};
|
||||
|
||||
// DELIBERATE DUPLICATION of the x-amz-checksum-* names that also exist as
|
||||
// AMZ_CHECKSUM_* in rustfs-utils' headers module (crates/utils/src/http/
|
||||
// headers.rs): this crate is a zero-internal-dependency leaf, so it cannot
|
||||
// import them, and it additionally owns the RustFS extension names
|
||||
// (sha512/xxhash*) that utils does not carry. Values are pinned by the S3
|
||||
// wire protocol; do not merge without a maintainer decision on the leaf
|
||||
// boundary (backlog#1833).
|
||||
pub const CRC_32_HEADER_NAME: &str = "x-amz-checksum-crc32";
|
||||
pub const CRC_32_C_HEADER_NAME: &str = "x-amz-checksum-crc32c";
|
||||
pub const SHA_1_HEADER_NAME: &str = "x-amz-checksum-sha1";
|
||||
|
||||
@@ -41,14 +41,6 @@ pub const XXHASH_64_NAME: &str = "xxhash64";
|
||||
pub const XXHASH_128_NAME: &str = "xxhash128";
|
||||
pub const MD5_NAME: &str = "md5";
|
||||
|
||||
/// One of three deliberately separate checksum registries (backlog#1833):
|
||||
/// this enum owns the **streaming-hash algorithm registry**, including the
|
||||
/// RustFS extensions (sha512, xxhash3/64/128). The on-disk xl.meta bitset
|
||||
/// lives in `rustfs_rio::ChecksumType` (crates/rio/src/checksum.rs, varint
|
||||
/// bits are append-only), and the MinIO-port client keeps its own
|
||||
/// `ChecksumMode` (crates/ecstore/src/client/checksum.rs). When adding an
|
||||
/// algorithm, extend all three (or record why not) — they do not derive from
|
||||
/// each other.
|
||||
#[derive(Debug, Clone, Copy, PartialEq, Eq, Default)]
|
||||
#[non_exhaustive]
|
||||
pub enum ChecksumAlgorithm {
|
||||
|
||||
@@ -0,0 +1,87 @@
|
||||
// Copyright 2024 RustFS Team
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
use crate::last_minute::{self};
|
||||
use std::collections::HashMap;
|
||||
|
||||
pub struct ReplicationLatency {
|
||||
// Delays for single and multipart PUT requests
|
||||
upload_histogram: last_minute::LastMinuteHistogram,
|
||||
}
|
||||
|
||||
impl ReplicationLatency {
|
||||
// Merge two ReplicationLatency
|
||||
pub fn merge(&mut self, other: &mut ReplicationLatency) -> &ReplicationLatency {
|
||||
self.upload_histogram.merge(&other.upload_histogram);
|
||||
self
|
||||
}
|
||||
|
||||
// Get upload delay (categorized by object size interval)
|
||||
pub fn get_upload_latency(&mut self) -> HashMap<String, u64> {
|
||||
let mut ret = HashMap::new();
|
||||
let avg = self.upload_histogram.get_avg_data();
|
||||
for (i, v) in avg.iter().enumerate() {
|
||||
let avg_duration = v.avg();
|
||||
ret.insert(self.size_tag_to_string(i), avg_duration.as_millis() as u64);
|
||||
}
|
||||
ret
|
||||
}
|
||||
pub fn update(&mut self, size: i64, during: std::time::Duration) {
|
||||
self.upload_histogram.add(size, during);
|
||||
}
|
||||
|
||||
// Simulate the conversion from size tag to string
|
||||
fn size_tag_to_string(&self, tag: usize) -> String {
|
||||
match tag {
|
||||
0 => String::from("Size < 1 KiB"),
|
||||
1 => String::from("Size < 1 MiB"),
|
||||
2 => String::from("Size < 10 MiB"),
|
||||
3 => String::from("Size < 100 MiB"),
|
||||
4 => String::from("Size < 1 GiB"),
|
||||
_ => String::from("Size > 1 GiB"),
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// #[derive(Debug, Clone, Default)]
|
||||
// pub struct ReplicationLastMinute {
|
||||
// pub last_minute: LastMinuteLatency,
|
||||
// }
|
||||
|
||||
// impl ReplicationLastMinute {
|
||||
// pub fn merge(&mut self, other: ReplicationLastMinute) -> ReplicationLastMinute {
|
||||
// let mut nl = ReplicationLastMinute::default();
|
||||
// nl.last_minute = self.last_minute.merge(&mut other.last_minute);
|
||||
// nl
|
||||
// }
|
||||
|
||||
// pub fn add_size(&mut self, n: i64) {
|
||||
// let t = SystemTime::now()
|
||||
// .duration_since(UNIX_EPOCH)
|
||||
// .expect("Time went backwards")
|
||||
// .as_secs();
|
||||
// self.last_minute.add_all(t - 1, &AccElem { total: t - 1, size: n as u64, n: 1 });
|
||||
// }
|
||||
|
||||
// pub fn get_total(&self) -> AccElem {
|
||||
// self.last_minute.get_total()
|
||||
// }
|
||||
// }
|
||||
|
||||
// impl fmt::Display for ReplicationLastMinute {
|
||||
// fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
|
||||
// let t = self.last_minute.get_total();
|
||||
// write!(f, "ReplicationLastMinute sz= {}, n= {}, dur= {}", t.size, t.n, t.total)
|
||||
// }
|
||||
// }
|
||||
@@ -572,3 +572,44 @@ mod tests {
|
||||
assert_eq!(total.n, 6);
|
||||
}
|
||||
}
|
||||
|
||||
const SIZE_LAST_ELEM_MARKER: usize = 10; // Assumed marker size is 10, modify according to actual situation
|
||||
|
||||
#[allow(dead_code)]
|
||||
#[derive(Debug, Default)]
|
||||
pub struct LastMinuteHistogram {
|
||||
histogram: Vec<LastMinuteLatency>,
|
||||
size: u32,
|
||||
}
|
||||
|
||||
impl LastMinuteHistogram {
|
||||
pub fn merge(&mut self, other: &LastMinuteHistogram) {
|
||||
for i in 0..self.histogram.len() {
|
||||
self.histogram[i].merge(&other.histogram[i]);
|
||||
}
|
||||
}
|
||||
|
||||
pub fn add(&mut self, size: i64, t: Duration) {
|
||||
let index = size_to_tag(size);
|
||||
self.histogram[index].add(&t);
|
||||
}
|
||||
|
||||
pub fn get_avg_data(&mut self) -> [AccElem; SIZE_LAST_ELEM_MARKER] {
|
||||
let mut res = [AccElem::default(); SIZE_LAST_ELEM_MARKER];
|
||||
for (i, elem) in self.histogram.iter_mut().enumerate() {
|
||||
res[i] = elem.get_total();
|
||||
}
|
||||
res
|
||||
}
|
||||
}
|
||||
|
||||
fn size_to_tag(size: i64) -> usize {
|
||||
match size {
|
||||
_ if size < 1024 => 0, // sizeLessThan1KiB
|
||||
_ if size < 1024 * 1024 => 1, // sizeLessThan1MiB
|
||||
_ if size < 10 * 1024 * 1024 => 2, // sizeLessThan10MiB
|
||||
_ if size < 100 * 1024 * 1024 => 3, // sizeLessThan100MiB
|
||||
_ if size < 1024 * 1024 * 1024 => 4, // sizeLessThan1GiB
|
||||
_ => 5, // sizeGreaterThan1GiB
|
||||
}
|
||||
}
|
||||
|
||||
@@ -12,6 +12,7 @@
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
pub mod bucket_stats;
|
||||
// pub mod error;
|
||||
pub mod globals;
|
||||
pub mod heal_channel;
|
||||
|
||||
@@ -353,11 +353,6 @@ pub const DEFAULT_OBS_TRACES_EXPORT_ENABLED: bool = true;
|
||||
/// Environment variable: RUSTFS_OBS_METRICS_EXPORT_ENABLED
|
||||
pub const DEFAULT_OBS_METRICS_EXPORT_ENABLED: bool = true;
|
||||
|
||||
/// Default detailed PUT stage metrics enabled
|
||||
/// Default value: false
|
||||
/// Environment variable: RUSTFS_OBS_PUT_STAGE_METRICS_ENABLED
|
||||
pub const DEFAULT_OBS_PUT_STAGE_METRICS_ENABLED: bool = false;
|
||||
|
||||
/// Default logs export enabled
|
||||
/// It is used to enable or disable exporting logs
|
||||
/// Default value: true
|
||||
|
||||
@@ -137,37 +137,6 @@ pub const DEFAULT_TIER_REMOTE_VERSION_STATE_FLEET_CONFIRMED: bool = false;
|
||||
const _: () = assert!(!DEFAULT_TIER_REMOTE_VERSION_STATE_WRITE);
|
||||
const _: () = assert!(!DEFAULT_TIER_REMOTE_VERSION_STATE_FLEET_CONFIRMED);
|
||||
|
||||
/// Request the object-transaction fencing contract used by storage-owned
|
||||
/// cleanup receipts and lock-window optimizations.
|
||||
///
|
||||
/// This is fail-closed: enabling the writer without a live fleet proof rejects
|
||||
/// the commit rather than silently using a legacy-safe path.
|
||||
pub const ENV_OBJECT_TRANSACTION_FENCING_WRITE: &str = "RUSTFS_OBJECT_TRANSACTION_FENCING_WRITE";
|
||||
pub const DEFAULT_OBJECT_TRANSACTION_FENCING_WRITE: bool = false;
|
||||
|
||||
/// Operator-attested confirmation that every serving node understands the
|
||||
/// object transaction fencing contract.
|
||||
pub const ENV_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED: &str = "RUSTFS_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED";
|
||||
pub const DEFAULT_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED: bool = false;
|
||||
|
||||
const _: () = assert!(!DEFAULT_OBJECT_TRANSACTION_FENCING_WRITE);
|
||||
const _: () = assert!(!DEFAULT_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED);
|
||||
|
||||
/// Request preserving legacy per-part checksum metadata during data movement.
|
||||
///
|
||||
/// This remains ineffective until
|
||||
/// [`ENV_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED`] is also enabled.
|
||||
pub const ENV_DATA_MOVEMENT_PART_CHECKSUMS_WRITE: &str = "RUSTFS_DATA_MOVEMENT_PART_CHECKSUMS_WRITE";
|
||||
pub const DEFAULT_DATA_MOVEMENT_PART_CHECKSUMS_WRITE: bool = false;
|
||||
|
||||
/// Operator-attested confirmation that every serving node understands the
|
||||
/// data-movement per-part checksum sidecar.
|
||||
pub const ENV_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED: &str = "RUSTFS_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED";
|
||||
pub const DEFAULT_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED: bool = false;
|
||||
|
||||
const _: () = assert!(!DEFAULT_DATA_MOVEMENT_PART_CHECKSUMS_WRITE);
|
||||
const _: () = assert!(!DEFAULT_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED);
|
||||
|
||||
// =============================================================================
|
||||
// Concurrent Request Fix - Timeout and Backpressure Configuration
|
||||
// =============================================================================
|
||||
@@ -680,22 +649,4 @@ mod remote_version_state_tests {
|
||||
"RUSTFS_TIER_REMOTE_VERSION_STATE_FLEET_CONFIRMED"
|
||||
);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn data_movement_part_checksum_gate_uses_stable_environment_names() {
|
||||
assert_eq!(super::ENV_DATA_MOVEMENT_PART_CHECKSUMS_WRITE, "RUSTFS_DATA_MOVEMENT_PART_CHECKSUMS_WRITE");
|
||||
assert_eq!(
|
||||
super::ENV_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED,
|
||||
"RUSTFS_DATA_MOVEMENT_PART_CHECKSUMS_FLEET_CONFIRMED"
|
||||
);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn object_transaction_fencing_gate_uses_stable_environment_names() {
|
||||
assert_eq!(super::ENV_OBJECT_TRANSACTION_FENCING_WRITE, "RUSTFS_OBJECT_TRANSACTION_FENCING_WRITE");
|
||||
assert_eq!(
|
||||
super::ENV_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED,
|
||||
"RUSTFS_OBJECT_TRANSACTION_FENCING_FLEET_CONFIRMED"
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
@@ -81,8 +81,7 @@ pub const ENV_TEST_IAM_FAIL_INIT_ATTEMPTS: &str = "RUSTFS_TEST_IAM_FAIL_INIT_ATT
|
||||
pub const ENV_TEST_IAM_RETRY_INTERVAL_MS: &str = "RUSTFS_TEST_IAM_RETRY_INTERVAL_MS";
|
||||
/// Runtime env var controlling the transition worker count.
|
||||
pub const ENV_TRANSITION_WORKERS: &str = "RUSTFS_MAX_TRANSITION_WORKERS";
|
||||
/// Runtime env var controlling the ILM expiry worker count. A set, parsable,
|
||||
/// non-zero value wins; anything else falls back to `min(cpus, 16)`.
|
||||
/// Runtime env var controlling the expiry worker count.
|
||||
pub const ENV_MAX_EXPIRY_WORKERS: &str = "RUSTFS_MAX_EXPIRY_WORKERS";
|
||||
/// Runtime env var controlling the absolute maximum transition workers.
|
||||
pub const ENV_TRANSITION_WORKERS_ABSOLUTE_MAX: &str = "RUSTFS_ABSOLUTE_MAX_WORKERS";
|
||||
|
||||
@@ -44,10 +44,6 @@ pub const ENV_OBS_METRICS_EXPORT_ENABLED: &str = "RUSTFS_OBS_METRICS_EXPORT_ENAB
|
||||
pub const ENV_OBS_LOGS_EXPORT_ENABLED: &str = "RUSTFS_OBS_LOGS_EXPORT_ENABLED";
|
||||
pub const ENV_OBS_PROFILING_EXPORT_ENABLED: &str = "RUSTFS_OBS_PROFILING_EXPORT_ENABLED";
|
||||
|
||||
/// Enables detailed per-stage PUT metrics. Disabled by default because each
|
||||
/// PUT records multiple timers and histograms when attribution is active.
|
||||
pub const ENV_OBS_PUT_STAGE_METRICS_ENABLED: &str = "RUSTFS_OBS_PUT_STAGE_METRICS_ENABLED";
|
||||
|
||||
pub const ENV_OBS_LOGGER_LEVEL: &str = "RUSTFS_OBS_LOGGER_LEVEL";
|
||||
pub const ENV_OBS_LOG_STDOUT_ENABLED: &str = "RUSTFS_OBS_LOG_STDOUT_ENABLED";
|
||||
pub const ENV_OBS_LOG_DIRECTORY: &str = "RUSTFS_OBS_LOG_DIRECTORY";
|
||||
@@ -145,7 +141,6 @@ mod tests {
|
||||
assert_eq!(ENV_OBS_METRICS_EXPORT_ENABLED, "RUSTFS_OBS_METRICS_EXPORT_ENABLED");
|
||||
assert_eq!(ENV_OBS_LOGS_EXPORT_ENABLED, "RUSTFS_OBS_LOGS_EXPORT_ENABLED");
|
||||
assert_eq!(ENV_OBS_PROFILING_EXPORT_ENABLED, "RUSTFS_OBS_PROFILING_EXPORT_ENABLED");
|
||||
assert_eq!(ENV_OBS_PUT_STAGE_METRICS_ENABLED, "RUSTFS_OBS_PUT_STAGE_METRICS_ENABLED");
|
||||
// Test log cleanup related env keys
|
||||
assert_eq!(ENV_OBS_LOG_MAX_TOTAL_SIZE_BYTES, "RUSTFS_OBS_LOG_MAX_TOTAL_SIZE_BYTES");
|
||||
assert_eq!(ENV_OBS_LOG_MAX_SINGLE_FILE_SIZE_BYTES, "RUSTFS_OBS_LOG_MAX_SINGLE_FILE_SIZE_BYTES");
|
||||
|
||||
@@ -67,10 +67,7 @@ fn configured_capture_log_path(temp_dir: &str) -> Option<String> {
|
||||
capture_log_path(Path::new(&log_dir), temp_dir).map(|path| path.to_string_lossy().into_owned())
|
||||
}
|
||||
|
||||
pub(crate) fn capture_command_logs(
|
||||
command: &mut Command,
|
||||
log_path: Option<&str>,
|
||||
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
fn capture_command_logs(command: &mut Command, log_path: Option<&str>) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
let Some(log_path) = log_path else {
|
||||
return Ok(());
|
||||
};
|
||||
|
||||
@@ -2,7 +2,6 @@
|
||||
|
||||
use crate::common::{RustFSTestEnvironment, init_logging, rustfs_binary_path};
|
||||
use aws_sdk_s3::primitives::ByteStream;
|
||||
use aws_sdk_s3::types::{CompletedMultipartUpload, CompletedPart};
|
||||
use serial_test::serial;
|
||||
use std::fs;
|
||||
use std::path::PathBuf;
|
||||
@@ -26,15 +25,6 @@ fn generate_compressible_data(size: usize) -> Vec<u8> {
|
||||
data
|
||||
}
|
||||
|
||||
/// Deterministic 2048-byte-period binary pattern that compresses extremely well: every part
|
||||
/// yields many compressed blocks, which is exactly the shape that reproduced the mid-payload
|
||||
/// Pending truncation (rustfs/rustfs#5957).
|
||||
fn generate_high_ratio_binary_data(size: usize, seed: u8) -> Vec<u8> {
|
||||
(0..size)
|
||||
.map(|i| ((i as u64).wrapping_mul(2_654_435_761).wrapping_add(seed as u64) >> 3) as u8)
|
||||
.collect()
|
||||
}
|
||||
|
||||
fn find_part_files(temp_dir: &str, bucket: &str, object_key: &str) -> Vec<PathBuf> {
|
||||
let bucket_path = PathBuf::from(temp_dir).join(bucket);
|
||||
let mut part_files = Vec::new();
|
||||
@@ -65,14 +55,9 @@ async fn start_rustfs_with_compression(env: &mut RustFSTestEnvironment) -> Resul
|
||||
env.cleanup_existing_processes().await?;
|
||||
|
||||
let binary_path = rustfs_binary_path();
|
||||
// Route the child's stdout/stderr through the shared RUSTFS_E2E_LOG_DIR
|
||||
// capture (survives the temp-dir cleanup on Drop and is uploaded as a CI
|
||||
// artifact); without the env var the child inherits stdio as before.
|
||||
let mut command = Command::new(&binary_path);
|
||||
command
|
||||
let process = Command::new(&binary_path)
|
||||
.env("RUSTFS_CONSOLE_ENABLE", "false")
|
||||
.env("RUSTFS_COMPRESSION_ENABLED", "true")
|
||||
.env("RUSTFS_COMPRESSION_MULTIPART_ENABLED", "true")
|
||||
.args([
|
||||
"--address",
|
||||
&env.address,
|
||||
@@ -81,9 +66,8 @@ async fn start_rustfs_with_compression(env: &mut RustFSTestEnvironment) -> Resul
|
||||
"--secret-key",
|
||||
&env.secret_key,
|
||||
&env.temp_dir,
|
||||
]);
|
||||
crate::common::capture_command_logs(&mut command, env.capture_log_path.as_deref())?;
|
||||
let process = command.spawn()?;
|
||||
])
|
||||
.spawn()?;
|
||||
|
||||
env.process = Some(process);
|
||||
|
||||
@@ -170,647 +154,3 @@ async fn test_compression_roundtrip() -> Result<(), Box<dyn std::error::Error +
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
const MULTIPART_COMPRESSION_BUCKET: &str = "compression-multipart-bucket";
|
||||
const MPU_PART1_SIZE: usize = 5 * 1024 * 1024;
|
||||
const MPU_PART2_SIZE: usize = 1024 * 1024;
|
||||
|
||||
async fn multipart_upload(
|
||||
client: &aws_sdk_s3::Client,
|
||||
bucket: &str,
|
||||
key: &str,
|
||||
parts: &[&[u8]],
|
||||
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
let create = client.create_multipart_upload().bucket(bucket).key(key).send().await?;
|
||||
let upload_id = create.upload_id().ok_or("missing upload id")?.to_string();
|
||||
|
||||
let mut completed_parts = Vec::with_capacity(parts.len());
|
||||
for (i, part) in parts.iter().enumerate() {
|
||||
let part_number = (i + 1) as i32;
|
||||
let upload = client
|
||||
.upload_part()
|
||||
.bucket(bucket)
|
||||
.key(key)
|
||||
.upload_id(&upload_id)
|
||||
.part_number(part_number)
|
||||
.body(ByteStream::from(part.to_vec()))
|
||||
.send()
|
||||
.await?;
|
||||
completed_parts.push(
|
||||
CompletedPart::builder()
|
||||
.part_number(part_number)
|
||||
.e_tag(upload.e_tag().unwrap_or_default())
|
||||
.build(),
|
||||
);
|
||||
}
|
||||
|
||||
client
|
||||
.complete_multipart_upload()
|
||||
.bucket(bucket)
|
||||
.key(key)
|
||||
.upload_id(&upload_id)
|
||||
.multipart_upload(CompletedMultipartUpload::builder().set_parts(Some(completed_parts)).build())
|
||||
.send()
|
||||
.await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
async fn fetch_range(
|
||||
client: &aws_sdk_s3::Client,
|
||||
bucket: &str,
|
||||
key: &str,
|
||||
range: &str,
|
||||
) -> Result<Vec<u8>, Box<dyn std::error::Error + Send + Sync>> {
|
||||
let response = client.get_object().bucket(bucket).key(key).range(range).send().await?;
|
||||
Ok(response.body.collect().await?.into_bytes().to_vec())
|
||||
}
|
||||
|
||||
/// Multipart disk compression roundtrip: parts are written as independent
|
||||
/// compressed streams and every GET shape must reassemble the original bytes
|
||||
/// (rustfs/rustfs#5957: multipart uploads previously bypassed disk compression
|
||||
/// entirely).
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn test_compression_multipart_roundtrip() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
init_logging();
|
||||
info!("Starting multipart compression roundtrip test");
|
||||
|
||||
let mut env = RustFSTestEnvironment::new().await?;
|
||||
start_rustfs_with_compression(&mut env).await?;
|
||||
|
||||
let client = env.create_s3_client();
|
||||
env.create_test_bucket(MULTIPART_COMPRESSION_BUCKET).await?;
|
||||
|
||||
let object_key = "multipart-compressible.txt";
|
||||
let part1 = generate_compressible_data(MPU_PART1_SIZE);
|
||||
let part2 = generate_compressible_data(MPU_PART2_SIZE);
|
||||
let mut original_data = part1.clone();
|
||||
original_data.extend_from_slice(&part2);
|
||||
let total_size = original_data.len();
|
||||
|
||||
multipart_upload(&client, MULTIPART_COMPRESSION_BUCKET, object_key, &[&part1, &part2]).await?;
|
||||
|
||||
let head_response = client
|
||||
.head_object()
|
||||
.bucket(MULTIPART_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
head_response.content_length().unwrap_or(0) as usize,
|
||||
total_size,
|
||||
"Content-Length should be the logical object size"
|
||||
);
|
||||
|
||||
let part_files = find_part_files(&env.temp_dir, MULTIPART_COMPRESSION_BUCKET, object_key);
|
||||
assert!(!part_files.is_empty(), "expected on-disk part files for the multipart object");
|
||||
let total_physical_size: u64 = part_files.iter().filter_map(|p| fs::metadata(p).ok()).map(|m| m.len()).sum();
|
||||
assert!(
|
||||
total_physical_size < (total_size / 2) as u64,
|
||||
"Physical size {total_physical_size} should be well below original size {total_size} (multipart compression applied)"
|
||||
);
|
||||
info!("Multipart physical storage size: {total_physical_size} bytes (compressed from {total_size} bytes)");
|
||||
|
||||
// Full GET must reassemble both independently compressed parts.
|
||||
let get_response = client
|
||||
.get_object()
|
||||
.bucket(MULTIPART_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
let downloaded = get_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(downloaded.len(), total_size);
|
||||
assert_eq!(&downloaded[..], &original_data[..], "full GET data mismatch");
|
||||
|
||||
// Range fully inside part 1.
|
||||
let range_inside_part1 = fetch_range(&client, MULTIPART_COMPRESSION_BUCKET, object_key, "bytes=1024-999423").await?;
|
||||
assert_eq!(&range_inside_part1[..], &original_data[1024..999424], "part-1 range mismatch");
|
||||
|
||||
// Range crossing the part boundary.
|
||||
let boundary_start = MPU_PART1_SIZE - 128 * 1024;
|
||||
let boundary_end = MPU_PART1_SIZE + 128 * 1024 - 1;
|
||||
let range_crossing = fetch_range(
|
||||
&client,
|
||||
MULTIPART_COMPRESSION_BUCKET,
|
||||
object_key,
|
||||
&format!("bytes={boundary_start}-{boundary_end}"),
|
||||
)
|
||||
.await?;
|
||||
assert_eq!(
|
||||
&range_crossing[..],
|
||||
&original_data[boundary_start..boundary_end + 1],
|
||||
"boundary-crossing range mismatch"
|
||||
);
|
||||
|
||||
// Range fully inside part 2.
|
||||
let part2_start = MPU_PART1_SIZE + 4096;
|
||||
let part2_end = MPU_PART1_SIZE + 256 * 1024 - 1;
|
||||
let range_inside_part2 = fetch_range(
|
||||
&client,
|
||||
MULTIPART_COMPRESSION_BUCKET,
|
||||
object_key,
|
||||
&format!("bytes={part2_start}-{part2_end}"),
|
||||
)
|
||||
.await?;
|
||||
assert_eq!(
|
||||
&range_inside_part2[..],
|
||||
&original_data[part2_start..part2_end + 1],
|
||||
"part-2 range mismatch"
|
||||
);
|
||||
|
||||
// Suffix range (last 128 KiB, entirely in part 2).
|
||||
let suffix_len = 128 * 1024;
|
||||
let suffix = fetch_range(&client, MULTIPART_COMPRESSION_BUCKET, object_key, &format!("bytes=-{suffix_len}")).await?;
|
||||
assert_eq!(&suffix[..], &original_data[total_size - suffix_len..], "suffix range mismatch");
|
||||
|
||||
// partNumber GETs must return each original part.
|
||||
for (part_number, expected) in [(1, &part1), (2, &part2)] {
|
||||
let response = client
|
||||
.get_object()
|
||||
.bucket(MULTIPART_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.part_number(part_number)
|
||||
.send()
|
||||
.await?;
|
||||
let body = response.body.collect().await?.into_bytes();
|
||||
assert_eq!(&body[..], &expected[..], "partNumber={part_number} GET mismatch");
|
||||
}
|
||||
|
||||
info!("Multipart compression roundtrip test passed");
|
||||
env.delete_test_bucket(MULTIPART_COMPRESSION_BUCKET).await?;
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
const MPU_HIGH_RATIO_BUCKET: &str = "compression-mpu-high-ratio-bucket";
|
||||
|
||||
/// High-ratio binary multipart payload: the object key is on the compression allow-list, so the
|
||||
/// disk-compression path runs and each part is stored as many compressed blocks — the shape that
|
||||
/// reproduced the mid-payload Pending truncation (rustfs/rustfs#5957). Every GET shape must return
|
||||
/// the exact original bytes, and the stored size must show the data really was compressed.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn test_compression_multipart_high_ratio_binary_roundtrip() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
init_logging();
|
||||
info!("Starting multipart high-ratio binary compression roundtrip test");
|
||||
|
||||
let mut env = RustFSTestEnvironment::new().await?;
|
||||
start_rustfs_with_compression(&mut env).await?;
|
||||
|
||||
let client = env.create_s3_client();
|
||||
env.create_test_bucket(MPU_HIGH_RATIO_BUCKET).await?;
|
||||
|
||||
let object_key = "multipart-high-ratio.txt";
|
||||
let part1 = generate_high_ratio_binary_data(MPU_PART1_SIZE, 7);
|
||||
let part2 = generate_high_ratio_binary_data(MPU_PART2_SIZE, 61);
|
||||
let mut original_data = part1.clone();
|
||||
original_data.extend_from_slice(&part2);
|
||||
let total_size = original_data.len();
|
||||
|
||||
multipart_upload(&client, MPU_HIGH_RATIO_BUCKET, object_key, &[&part1, &part2]).await?;
|
||||
|
||||
let head_response = client
|
||||
.head_object()
|
||||
.bucket(MPU_HIGH_RATIO_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
head_response.content_length().unwrap_or(0) as usize,
|
||||
total_size,
|
||||
"Content-Length should be the logical object size"
|
||||
);
|
||||
|
||||
// This pattern compresses to roughly 1/50 of its logical size, so a comfortably loose 2x
|
||||
// margin still proves the parts were stored compressed rather than raw or double-encoded.
|
||||
let part_files = find_part_files(&env.temp_dir, MPU_HIGH_RATIO_BUCKET, object_key);
|
||||
assert!(!part_files.is_empty(), "expected on-disk part files for the multipart object");
|
||||
let total_physical_size: u64 = part_files.iter().filter_map(|p| fs::metadata(p).ok()).map(|m| m.len()).sum();
|
||||
assert!(
|
||||
total_physical_size < (total_size as u64) / 2,
|
||||
"Physical size {total_physical_size} should be far below the logical size {total_size} for high-ratio data"
|
||||
);
|
||||
info!("High-ratio multipart physical storage size: {total_physical_size} bytes (logical {total_size} bytes)");
|
||||
|
||||
info!("step: full GET");
|
||||
let get_response = client
|
||||
.get_object()
|
||||
.bucket(MPU_HIGH_RATIO_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
let downloaded = get_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(downloaded.len(), total_size);
|
||||
assert_eq!(&downloaded[..], &original_data[..], "full GET data mismatch");
|
||||
|
||||
// Range crossing the part boundary.
|
||||
info!("step: boundary range GET");
|
||||
let boundary_start = MPU_PART1_SIZE - 128 * 1024;
|
||||
let boundary_end = MPU_PART1_SIZE + 128 * 1024 - 1;
|
||||
let range_crossing = fetch_range(
|
||||
&client,
|
||||
MPU_HIGH_RATIO_BUCKET,
|
||||
object_key,
|
||||
&format!("bytes={boundary_start}-{boundary_end}"),
|
||||
)
|
||||
.await?;
|
||||
assert_eq!(
|
||||
&range_crossing[..],
|
||||
&original_data[boundary_start..boundary_end + 1],
|
||||
"boundary-crossing range mismatch"
|
||||
);
|
||||
|
||||
// partNumber GET for the trailing part.
|
||||
info!("step: partNumber GET");
|
||||
let part2_response = client
|
||||
.get_object()
|
||||
.bucket(MPU_HIGH_RATIO_BUCKET)
|
||||
.key(object_key)
|
||||
.part_number(2)
|
||||
.send()
|
||||
.await?;
|
||||
let part2_body = part2_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(&part2_body[..], &part2[..], "partNumber=2 GET mismatch");
|
||||
|
||||
info!("Multipart high-ratio binary compression roundtrip test passed");
|
||||
env.delete_test_bucket(MPU_HIGH_RATIO_BUCKET).await?;
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
const MPU_COPY_COMPRESSION_BUCKET: &str = "compression-mpu-copy-bucket";
|
||||
const MPU_COPY_SOURCE_SIZE: usize = 6 * 1024 * 1024;
|
||||
const MPU_COPY_RANGE_LEN: usize = 5 * 1024 * 1024;
|
||||
|
||||
/// UploadPartCopy feeds a part from an already stored (and already compressed) object. The copied
|
||||
/// range must be decompressed on read and re-compressed into the destination part, so the final
|
||||
/// object has to match "source prefix + uploaded tail" byte for byte.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn test_compression_multipart_upload_part_copy_roundtrip() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
init_logging();
|
||||
info!("Starting multipart upload-part-copy compression roundtrip test");
|
||||
|
||||
let mut env = RustFSTestEnvironment::new().await?;
|
||||
start_rustfs_with_compression(&mut env).await?;
|
||||
|
||||
let client = env.create_s3_client();
|
||||
env.create_test_bucket(MPU_COPY_COMPRESSION_BUCKET).await?;
|
||||
|
||||
// Source object: a plain PUT that goes through the single-stream compression path.
|
||||
let source_key = "copy-source.txt";
|
||||
let source_data = generate_compressible_data(MPU_COPY_SOURCE_SIZE);
|
||||
client
|
||||
.put_object()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(source_key)
|
||||
.body(ByteStream::from(source_data.clone()))
|
||||
.send()
|
||||
.await?;
|
||||
|
||||
// Destination object: part 1 copied from the source, part 2 uploaded directly.
|
||||
let target_key = "copy-target.txt";
|
||||
let part2 = generate_compressible_data(MPU_PART2_SIZE);
|
||||
let mut expected_data = source_data[..MPU_COPY_RANGE_LEN].to_vec();
|
||||
expected_data.extend_from_slice(&part2);
|
||||
let total_size = expected_data.len();
|
||||
|
||||
let create = client
|
||||
.create_multipart_upload()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.send()
|
||||
.await?;
|
||||
let upload_id = create.upload_id().ok_or("missing upload id")?.to_string();
|
||||
|
||||
let copy_part = client
|
||||
.upload_part_copy()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.upload_id(&upload_id)
|
||||
.part_number(1)
|
||||
.copy_source(format!("{MPU_COPY_COMPRESSION_BUCKET}/{source_key}"))
|
||||
.copy_source_range(format!("bytes=0-{}", MPU_COPY_RANGE_LEN - 1))
|
||||
.send()
|
||||
.await?;
|
||||
let copy_etag = copy_part
|
||||
.copy_part_result()
|
||||
.and_then(|r| r.e_tag())
|
||||
.ok_or("missing copy part etag")?
|
||||
.to_string();
|
||||
|
||||
let uploaded_part = client
|
||||
.upload_part()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.upload_id(&upload_id)
|
||||
.part_number(2)
|
||||
.body(ByteStream::from(part2.clone()))
|
||||
.send()
|
||||
.await?;
|
||||
|
||||
client
|
||||
.complete_multipart_upload()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.upload_id(&upload_id)
|
||||
.multipart_upload(
|
||||
CompletedMultipartUpload::builder()
|
||||
.parts(CompletedPart::builder().part_number(1).e_tag(copy_etag).build())
|
||||
.parts(
|
||||
CompletedPart::builder()
|
||||
.part_number(2)
|
||||
.e_tag(uploaded_part.e_tag().unwrap_or_default())
|
||||
.build(),
|
||||
)
|
||||
.build(),
|
||||
)
|
||||
.send()
|
||||
.await?;
|
||||
|
||||
let head_response = client
|
||||
.head_object()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
head_response.content_length().unwrap_or(0) as usize,
|
||||
total_size,
|
||||
"Content-Length should be the logical object size"
|
||||
);
|
||||
|
||||
let part_files = find_part_files(&env.temp_dir, MPU_COPY_COMPRESSION_BUCKET, target_key);
|
||||
assert!(!part_files.is_empty(), "expected on-disk part files for the copied object");
|
||||
let total_physical_size: u64 = part_files.iter().filter_map(|p| fs::metadata(p).ok()).map(|m| m.len()).sum();
|
||||
assert!(
|
||||
total_physical_size < (total_size / 2) as u64,
|
||||
"Physical size {total_physical_size} should be well below original size {total_size} (copied part compression applied)"
|
||||
);
|
||||
|
||||
let get_response = client
|
||||
.get_object()
|
||||
.bucket(MPU_COPY_COMPRESSION_BUCKET)
|
||||
.key(target_key)
|
||||
.send()
|
||||
.await?;
|
||||
let downloaded = get_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(downloaded.len(), total_size);
|
||||
assert_eq!(&downloaded[..], &expected_data[..], "copied multipart GET data mismatch");
|
||||
|
||||
info!("Multipart upload-part-copy compression roundtrip test passed");
|
||||
env.delete_test_bucket(MPU_COPY_COMPRESSION_BUCKET).await?;
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
const MPU_THREE_PARTS_BUCKET: &str = "compression-mpu-three-parts-bucket";
|
||||
const MPU_THREE_PARTS_TAIL_SIZE: usize = 512 * 1024;
|
||||
|
||||
/// Three-part upload with uneven part sizes: each partNumber GET must map back to exactly one
|
||||
/// compressed part stream, and a suffix range must resolve inside the trailing part.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn test_compression_multipart_three_parts_part_number_gets() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
init_logging();
|
||||
info!("Starting three-part multipart compression partNumber test");
|
||||
|
||||
let mut env = RustFSTestEnvironment::new().await?;
|
||||
start_rustfs_with_compression(&mut env).await?;
|
||||
|
||||
let client = env.create_s3_client();
|
||||
env.create_test_bucket(MPU_THREE_PARTS_BUCKET).await?;
|
||||
|
||||
let object_key = "multipart-three-parts.txt";
|
||||
let part1 = generate_compressible_data(MPU_PART1_SIZE);
|
||||
let part2 = generate_compressible_data(MPU_PART1_SIZE);
|
||||
let part3 = generate_compressible_data(MPU_THREE_PARTS_TAIL_SIZE);
|
||||
let mut original_data = part1.clone();
|
||||
original_data.extend_from_slice(&part2);
|
||||
original_data.extend_from_slice(&part3);
|
||||
let total_size = original_data.len();
|
||||
|
||||
multipart_upload(&client, MPU_THREE_PARTS_BUCKET, object_key, &[&part1, &part2, &part3]).await?;
|
||||
|
||||
let head_response = client
|
||||
.head_object()
|
||||
.bucket(MPU_THREE_PARTS_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
head_response.content_length().unwrap_or(0) as usize,
|
||||
total_size,
|
||||
"Content-Length should be the logical object size"
|
||||
);
|
||||
|
||||
let part_files = find_part_files(&env.temp_dir, MPU_THREE_PARTS_BUCKET, object_key);
|
||||
assert!(!part_files.is_empty(), "expected on-disk part files for the multipart object");
|
||||
let total_physical_size: u64 = part_files.iter().filter_map(|p| fs::metadata(p).ok()).map(|m| m.len()).sum();
|
||||
assert!(
|
||||
total_physical_size < (total_size / 2) as u64,
|
||||
"Physical size {total_physical_size} should be well below original size {total_size} (multipart compression applied)"
|
||||
);
|
||||
|
||||
// Every partNumber GET must return exactly the bytes of the corresponding uploaded part.
|
||||
for (part_number, expected) in [(1, &part1), (2, &part2), (3, &part3)] {
|
||||
let response = client
|
||||
.get_object()
|
||||
.bucket(MPU_THREE_PARTS_BUCKET)
|
||||
.key(object_key)
|
||||
.part_number(part_number)
|
||||
.send()
|
||||
.await?;
|
||||
let body = response.body.collect().await?.into_bytes();
|
||||
assert_eq!(&body[..], &expected[..], "partNumber={part_number} GET mismatch");
|
||||
}
|
||||
|
||||
// Suffix range (last 64 KiB) resolves inside the trailing part.
|
||||
let suffix_len = 64 * 1024;
|
||||
let suffix = fetch_range(&client, MPU_THREE_PARTS_BUCKET, object_key, &format!("bytes=-{suffix_len}")).await?;
|
||||
assert_eq!(&suffix[..], &original_data[total_size - suffix_len..], "suffix range mismatch");
|
||||
|
||||
info!("Three-part multipart compression partNumber test passed");
|
||||
env.delete_test_bucket(MPU_THREE_PARTS_BUCKET).await?;
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
const MPU_SSE_COMPRESSION_BUCKET: &str = "compression-mpu-sse-bucket";
|
||||
|
||||
async fn start_rustfs_with_compression_and_sse(
|
||||
env: &mut RustFSTestEnvironment,
|
||||
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
use base64::Engine;
|
||||
env.cleanup_existing_processes().await?;
|
||||
|
||||
let binary_path = rustfs_binary_path();
|
||||
let master_key = base64::engine::general_purpose::STANDARD.encode([0x42u8; 32]);
|
||||
// Server output goes to a file inside the per-test temp dir so a failing
|
||||
// run can be diagnosed from the child's logs.
|
||||
let server_log = std::fs::File::create(format!("{}/server.log", env.temp_dir))?;
|
||||
let server_log_err = server_log.try_clone()?;
|
||||
let process = Command::new(&binary_path)
|
||||
.env("RUSTFS_CONSOLE_ENABLE", "false")
|
||||
.env("RUSTFS_COMPRESSION_ENABLED", "true")
|
||||
.env("RUSTFS_COMPRESSION_MULTIPART_ENABLED", "true")
|
||||
.env("RUSTFS_SSE_S3_MASTER_KEY", master_key)
|
||||
.env("RUST_LOG", "rustfs=info,rustfs_ecstore=info")
|
||||
.stdout(std::process::Stdio::from(server_log))
|
||||
.stderr(std::process::Stdio::from(server_log_err))
|
||||
.args([
|
||||
"--address",
|
||||
&env.address,
|
||||
"--access-key",
|
||||
&env.access_key,
|
||||
"--secret-key",
|
||||
&env.secret_key,
|
||||
&env.temp_dir,
|
||||
])
|
||||
.spawn()?;
|
||||
|
||||
env.process = Some(process);
|
||||
|
||||
info!("Waiting for RustFS server with compression + SSE-S3 enabled on {}", env.address);
|
||||
for i in 0..30 {
|
||||
if TcpStream::connect(&env.address).await.is_ok() {
|
||||
info!("RustFS server is ready after {} attempts", i + 1);
|
||||
return Ok(());
|
||||
}
|
||||
if i == 29 {
|
||||
return Err("RustFS server failed to become ready".into());
|
||||
}
|
||||
sleep(Duration::from_secs(1)).await;
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// SSE-S3 + disk compression multipart: each part is compressed and then encrypted, and every GET
|
||||
/// shape must still return the original plaintext bytes. Physical size must shrink because the
|
||||
/// compression runs before encryption.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn test_compression_multipart_sse_s3_roundtrip() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
use aws_sdk_s3::types::ServerSideEncryption;
|
||||
|
||||
init_logging();
|
||||
info!("Starting SSE-S3 multipart compression roundtrip test");
|
||||
|
||||
let mut env = RustFSTestEnvironment::new().await?;
|
||||
start_rustfs_with_compression_and_sse(&mut env).await?;
|
||||
|
||||
let client = env.create_s3_client();
|
||||
env.create_test_bucket(MPU_SSE_COMPRESSION_BUCKET).await?;
|
||||
|
||||
let object_key = "multipart-sse-compressible.txt";
|
||||
let part1 = generate_compressible_data(MPU_PART1_SIZE);
|
||||
let part2 = generate_compressible_data(MPU_PART2_SIZE);
|
||||
let mut original_data = part1.clone();
|
||||
original_data.extend_from_slice(&part2);
|
||||
let total_size = original_data.len();
|
||||
|
||||
let create = client
|
||||
.create_multipart_upload()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.server_side_encryption(ServerSideEncryption::Aes256)
|
||||
.send()
|
||||
.await?;
|
||||
let upload_id = create.upload_id().ok_or("missing upload id")?.to_string();
|
||||
|
||||
let mut completed_parts = Vec::new();
|
||||
for (i, part) in [&part1, &part2].into_iter().enumerate() {
|
||||
let part_number = (i + 1) as i32;
|
||||
let upload = client
|
||||
.upload_part()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.upload_id(&upload_id)
|
||||
.part_number(part_number)
|
||||
.body(ByteStream::from(part.clone()))
|
||||
.send()
|
||||
.await?;
|
||||
completed_parts.push(
|
||||
CompletedPart::builder()
|
||||
.part_number(part_number)
|
||||
.e_tag(upload.e_tag().unwrap_or_default())
|
||||
.build(),
|
||||
);
|
||||
}
|
||||
|
||||
client
|
||||
.complete_multipart_upload()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.upload_id(&upload_id)
|
||||
.multipart_upload(CompletedMultipartUpload::builder().set_parts(Some(completed_parts)).build())
|
||||
.send()
|
||||
.await?;
|
||||
|
||||
let head_response = client
|
||||
.head_object()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
head_response.content_length().unwrap_or(0) as usize,
|
||||
total_size,
|
||||
"Content-Length should be the logical object size"
|
||||
);
|
||||
assert_eq!(
|
||||
head_response.server_side_encryption(),
|
||||
Some(&ServerSideEncryption::Aes256),
|
||||
"HEAD must report SSE-S3"
|
||||
);
|
||||
|
||||
let part_files = find_part_files(&env.temp_dir, MPU_SSE_COMPRESSION_BUCKET, object_key);
|
||||
assert!(!part_files.is_empty(), "expected on-disk part files for the multipart object");
|
||||
let total_physical_size: u64 = part_files.iter().filter_map(|p| fs::metadata(p).ok()).map(|m| m.len()).sum();
|
||||
assert!(
|
||||
total_physical_size < (total_size / 2) as u64,
|
||||
"Physical size {total_physical_size} should be well below original size {total_size} (compress-then-encrypt applied)"
|
||||
);
|
||||
|
||||
let get_response = client
|
||||
.get_object()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.send()
|
||||
.await?;
|
||||
let downloaded = get_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(downloaded.len(), total_size);
|
||||
assert_eq!(&downloaded[..], &original_data[..], "SSE-S3 multipart full GET data mismatch");
|
||||
|
||||
// Range crossing the part boundary must decrypt and decompress across parts.
|
||||
let boundary_start = MPU_PART1_SIZE - 64 * 1024;
|
||||
let boundary_end = MPU_PART1_SIZE + 64 * 1024 - 1;
|
||||
let range_crossing = fetch_range(
|
||||
&client,
|
||||
MPU_SSE_COMPRESSION_BUCKET,
|
||||
object_key,
|
||||
&format!("bytes={boundary_start}-{boundary_end}"),
|
||||
)
|
||||
.await?;
|
||||
assert_eq!(
|
||||
&range_crossing[..],
|
||||
&original_data[boundary_start..boundary_end + 1],
|
||||
"SSE-S3 boundary-crossing range mismatch"
|
||||
);
|
||||
|
||||
// partNumber GET for the trailing part.
|
||||
let part2_response = client
|
||||
.get_object()
|
||||
.bucket(MPU_SSE_COMPRESSION_BUCKET)
|
||||
.key(object_key)
|
||||
.part_number(2)
|
||||
.send()
|
||||
.await?;
|
||||
let part2_body = part2_response.body.collect().await?.into_bytes();
|
||||
assert_eq!(&part2_body[..], &part2[..], "SSE-S3 partNumber=2 GET mismatch");
|
||||
|
||||
info!("SSE-S3 multipart compression roundtrip test passed");
|
||||
env.delete_test_bucket(MPU_SSE_COMPRESSION_BUCKET).await?;
|
||||
env.stop_server();
|
||||
Ok(())
|
||||
}
|
||||
|
||||
@@ -76,18 +76,6 @@ const SOURCE_MTIME_HEADERS: [&str; 2] = ["x-rustfs-source-mtime", "x-minio-sourc
|
||||
const SOURCE_REPLICATION_REQUEST_HEADERS: [&str; 2] =
|
||||
["x-rustfs-source-replication-request", "x-minio-source-replication-request"];
|
||||
const SOURCE_ETAG_HEADERS: [&str; 2] = ["x-rustfs-source-etag", "x-minio-source-etag"];
|
||||
const SOURCE_TAGGING_TIMESTAMP_HEADERS: [&str; 2] = [
|
||||
"x-rustfs-source-replication-tagging-timestamp",
|
||||
"x-minio-source-replication-tagging-timestamp",
|
||||
];
|
||||
const SOURCE_RETENTION_TIMESTAMP_HEADERS: [&str; 2] = [
|
||||
"x-rustfs-source-replication-retention-timestamp",
|
||||
"x-minio-source-replication-retention-timestamp",
|
||||
];
|
||||
const SOURCE_LEGALHOLD_TIMESTAMP_HEADERS: [&str; 2] = [
|
||||
"x-rustfs-source-replication-legalhold-timestamp",
|
||||
"x-minio-source-replication-legalhold-timestamp",
|
||||
];
|
||||
const RESERVED_BUCKET_PREFIXES: [&str; 3] = ["xn--", "sthree-", "amzn-s3-demo-"];
|
||||
const RESERVED_BUCKET_SUFFIXES: [&str; 6] = ["-s3alias", "--ol-s3", ".mrap", "--x-s3", "--table-s3", "-an"];
|
||||
|
||||
@@ -130,25 +118,6 @@ pub enum FaultAction {
|
||||
WrongEtag,
|
||||
}
|
||||
|
||||
/// Replication LWW timestamp headers observed on a request, journaled so
|
||||
/// sender-side tests can assert what a real target would receive.
|
||||
#[derive(Debug, Clone, Default, PartialEq, Eq)]
|
||||
pub struct ReplicationTimestampHeaders {
|
||||
pub tagging: Option<String>,
|
||||
pub retention: Option<String>,
|
||||
pub legalhold: Option<String>,
|
||||
}
|
||||
|
||||
impl ReplicationTimestampHeaders {
|
||||
fn from_headers(headers: &HeaderMap) -> Self {
|
||||
Self {
|
||||
tagging: header_value(headers, &SOURCE_TAGGING_TIMESTAMP_HEADERS).map(bounded_journal_value),
|
||||
retention: header_value(headers, &SOURCE_RETENTION_TIMESTAMP_HEADERS).map(bounded_journal_value),
|
||||
legalhold: header_value(headers, &SOURCE_LEGALHOLD_TIMESTAMP_HEADERS).map(bounded_journal_value),
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/// Credential-free request metadata retained for deterministic assertions.
|
||||
#[derive(Debug, Clone, PartialEq, Eq)]
|
||||
pub struct RequestRecord {
|
||||
@@ -162,7 +131,6 @@ pub struct RequestRecord {
|
||||
pub part_number: Option<i32>,
|
||||
pub content_length: Option<u64>,
|
||||
pub consumed_bytes: Option<usize>,
|
||||
pub replication_timestamps: ReplicationTimestampHeaders,
|
||||
pub fault: Option<FaultAction>,
|
||||
}
|
||||
|
||||
@@ -568,15 +536,7 @@ impl S3Access for FaultAccess {
|
||||
.get(CONTENT_LENGTH)
|
||||
.and_then(|value| value.to_str().ok())
|
||||
.and_then(|value| value.parse().ok());
|
||||
let replication_timestamps = ReplicationTimestampHeaders::from_headers(context.headers());
|
||||
let fault = record_request(
|
||||
&self.control,
|
||||
operation,
|
||||
context.method().clone(),
|
||||
parsed,
|
||||
content_length,
|
||||
replication_timestamps,
|
||||
);
|
||||
let fault = record_request(&self.control, operation, context.method().clone(), parsed, content_length);
|
||||
if let Some(RequestFault {
|
||||
action: FaultAction::Status(status),
|
||||
..
|
||||
@@ -629,7 +589,6 @@ fn record_request(
|
||||
method: Method,
|
||||
parsed: ParsedRequest,
|
||||
content_length: Option<u64>,
|
||||
replication_timestamps: ReplicationTimestampHeaders,
|
||||
) -> Option<RequestFault> {
|
||||
let mut state = lock(control);
|
||||
let action = parsed
|
||||
@@ -654,7 +613,6 @@ fn record_request(
|
||||
part_number: parsed.part_number,
|
||||
content_length,
|
||||
consumed_bytes: None,
|
||||
replication_timestamps,
|
||||
fault: action.clone(),
|
||||
});
|
||||
action.map(|action| RequestFault { sequence, action })
|
||||
@@ -1741,52 +1699,6 @@ mod tests {
|
||||
.await?)
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn journals_replication_timestamp_headers() -> Result<(), BoxError> {
|
||||
let target = FakeS3Target::start().await?;
|
||||
target.create_bucket("target-bucket");
|
||||
let client = client(&target);
|
||||
|
||||
client
|
||||
.put_object()
|
||||
.bucket("target-bucket")
|
||||
.key("plain")
|
||||
.body(ByteStream::from_static(b"plain"))
|
||||
.send()
|
||||
.await?;
|
||||
client
|
||||
.put_object()
|
||||
.bucket("target-bucket")
|
||||
.key("stamped")
|
||||
.body(ByteStream::from_static(b"stamped"))
|
||||
.customize()
|
||||
.map_request(move |mut request| {
|
||||
let headers = request.headers_mut();
|
||||
headers.insert("x-rustfs-source-replication-tagging-timestamp", "2026-01-02T03:04:05Z");
|
||||
headers.insert("x-minio-source-replication-retention-timestamp", "2026-01-02T03:04:06Z");
|
||||
headers.insert("x-rustfs-source-replication-legalhold-timestamp", "2026-01-02T03:04:07Z");
|
||||
Ok::<_, std::convert::Infallible>(request)
|
||||
})
|
||||
.send()
|
||||
.await?;
|
||||
|
||||
let requests = target.requests();
|
||||
let plain = requests
|
||||
.iter()
|
||||
.find(|record| record.operation == Operation::PutObject && record.key.as_deref() == Some("plain"))
|
||||
.expect("plain PUT must be journaled");
|
||||
assert_eq!(plain.replication_timestamps, ReplicationTimestampHeaders::default());
|
||||
|
||||
let stamped = requests
|
||||
.iter()
|
||||
.find(|record| record.operation == Operation::PutObject && record.key.as_deref() == Some("stamped"))
|
||||
.expect("stamped PUT must be journaled");
|
||||
assert_eq!(stamped.replication_timestamps.tagging.as_deref(), Some("2026-01-02T03:04:05Z"));
|
||||
assert_eq!(stamped.replication_timestamps.retention.as_deref(), Some("2026-01-02T03:04:06Z"));
|
||||
assert_eq!(stamped.replication_timestamps.legalhold.as_deref(), Some("2026-01-02T03:04:07Z"));
|
||||
Ok(())
|
||||
}
|
||||
|
||||
macro_rules! assert_sdk_error {
|
||||
($error:expr, $status:expr, $code:expr) => {{
|
||||
let error = &$error;
|
||||
@@ -3073,7 +2985,6 @@ mod tests {
|
||||
part_number: None,
|
||||
},
|
||||
Some(0),
|
||||
ReplicationTimestampHeaders::default(),
|
||||
);
|
||||
}
|
||||
let records = lock(&control).requests.clone();
|
||||
@@ -3095,7 +3006,6 @@ mod tests {
|
||||
part_number: None,
|
||||
},
|
||||
None,
|
||||
ReplicationTimestampHeaders::default(),
|
||||
);
|
||||
{
|
||||
let bounded_records = lock(&bounded_control);
|
||||
|
||||
@@ -1828,36 +1828,33 @@ async fn four_node_compressed_inline_fallback() -> TestResult {
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Multipart disk compression is live again, so a compression-enabled cluster classifies multipart objects as compressed and the roundtrip (full GET plus partNumber GET) must still return the original bytes.
|
||||
/// Reverting the multipart compression fix must fail this test.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn four_node_multipart_disk_compression_roundtrip() -> TestResult {
|
||||
async fn four_node_multipart_ignores_disk_compression_fallback() -> TestResult {
|
||||
init_logging();
|
||||
|
||||
let collector = OtlpMetricCollector::start().await?;
|
||||
let mut cluster = RustFSTestClusterEnvironment::new(4).await?;
|
||||
configure_reader_metric_cluster(&mut cluster, &collector);
|
||||
cluster.set_env("RUSTFS_COMPRESSION_ENABLED", "true");
|
||||
cluster.set_env("RUSTFS_COMPRESSION_MULTIPART_ENABLED", "true");
|
||||
cluster.start().await?;
|
||||
|
||||
let bucket = "inline-multipart-compression-roundtrip";
|
||||
let bucket = "inline-multipart-compression-fallback";
|
||||
cluster.create_test_bucket(bucket).await?;
|
||||
let client = cluster.create_s3_client(0)?;
|
||||
let key = "multipart/compressed.txt";
|
||||
let key = "multipart/compression-disabled.txt";
|
||||
let (body, second_part, etag) = put_two_part_multipart(&client, bucket, key).await?;
|
||||
|
||||
assert_reader_path(
|
||||
&collector,
|
||||
&client,
|
||||
ReaderPathExpectation::for_class(ReaderObject::new(bucket, key, &body, etag.as_deref(), None), LEGACY_DUPLEX, COMPRESSED),
|
||||
ReaderPathExpectation::for_class(ReaderObject::new(bucket, key, &body, etag.as_deref(), None), LEGACY_DUPLEX, MULTIPART),
|
||||
)
|
||||
.await?;
|
||||
assert_part_number_reader_path(
|
||||
&collector,
|
||||
&client,
|
||||
PartNumberReaderPathExpectation::new(bucket, key, &second_part, body.len(), COMPRESSED, LEGACY_DUPLEX),
|
||||
PartNumberReaderPathExpectation::new(bucket, key, &second_part, body.len(), MULTIPART, LEGACY_DUPLEX),
|
||||
)
|
||||
.await?;
|
||||
|
||||
@@ -1874,7 +1871,6 @@ async fn four_node_mixed_msgpack_compat_mode_preserves_fallback_controls() -> Te
|
||||
let sse_master_key = base64::engine::general_purpose::STANDARD.encode([0x42u8; 32]);
|
||||
cluster.set_env("RUSTFS_SSE_S3_MASTER_KEY", sse_master_key);
|
||||
cluster.set_env("RUSTFS_COMPRESSION_ENABLED", "true");
|
||||
cluster.set_env("RUSTFS_COMPRESSION_MULTIPART_ENABLED", "true");
|
||||
configure_mixed_msgpack_cluster(&mut cluster, &collector)?;
|
||||
cluster.start().await?;
|
||||
|
||||
@@ -1894,21 +1890,14 @@ async fn four_node_mixed_msgpack_compat_mode_preserves_fallback_controls() -> Te
|
||||
ReaderPathExpectation::for_class(
|
||||
ReaderObject::new(bucket, multipart_key, &multipart_body, multipart_etag.as_deref(), None),
|
||||
LEGACY_DUPLEX,
|
||||
COMPRESSED,
|
||||
MULTIPART,
|
||||
),
|
||||
)
|
||||
.await?;
|
||||
assert_part_number_reader_path(
|
||||
&collector,
|
||||
&client,
|
||||
PartNumberReaderPathExpectation::new(
|
||||
bucket,
|
||||
multipart_key,
|
||||
&second_part,
|
||||
multipart_body.len(),
|
||||
COMPRESSED,
|
||||
LEGACY_DUPLEX,
|
||||
),
|
||||
PartNumberReaderPathExpectation::new(bucket, multipart_key, &second_part, multipart_body.len(), MULTIPART, LEGACY_DUPLEX),
|
||||
)
|
||||
.await?;
|
||||
assert_msgpack_decode_observed(&collector, &decode_before).await?;
|
||||
@@ -2364,11 +2353,7 @@ async fn four_node_mixed_msgpack_compat_mode_preserves_fallback_controls_during_
|
||||
hot_client.create_bucket().bucket(bucket).send().await?;
|
||||
put_lifecycle_with_transition_retry(&hot_client, bucket, &tier_name).await?;
|
||||
|
||||
// `.zip` sits on the disk-compression exclusion list: this test pins
|
||||
// msgpack compat controls across ILM transition, and a compressed object
|
||||
// would classify as `compressed` instead of `remote` (and the warm-tier
|
||||
// read path does not decode compression — tracked separately).
|
||||
let key = "transition/mixed-multipart.zip";
|
||||
let key = "transition/mixed-multipart.bin";
|
||||
let (body, second_part, etag) = put_two_part_multipart(&hot_client, bucket, key).await?;
|
||||
wait_for_transition(&hot_client, bucket, key, &tier_name).await?;
|
||||
assert!(
|
||||
|
||||
@@ -1,611 +0,0 @@
|
||||
// Copyright 2024 RustFS Team
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
//! ILM on SSE-KMS buckets while per-key SSE authorization is enforced (backlog#1582).
|
||||
//!
|
||||
//! Per-key KMS authorization (`RUSTFS_KMS_ENFORCE_SSE_KEY_POLICY=true`) scopes the
|
||||
//! SSE-KMS data path to the requesting principal's `kms:GenerateDataKey` /
|
||||
//! `kms:Decrypt` grants. Internal callers — the lifecycle scanner's expiry deletes
|
||||
//! and the tier transition worker's reads — carry no request principal, and
|
||||
//! `authorize_sse_kms_key` (rustfs/src/storage/sse.rs) exempts a `None` principal
|
||||
//! so background maintenance keeps working on encrypted buckets.
|
||||
//!
|
||||
//! These tests pin that exemption end to end. If enforcement ever starts applying
|
||||
//! to the scanner's internal operations, expiry stops happening on SSE-KMS buckets
|
||||
//! and [`ilm_expiration_on_sse_kms_bucket_under_enforcement`] times out; if it
|
||||
//! starts applying to the transition worker or the read-through path,
|
||||
//! [`ilm_transition_on_sse_kms_bucket_under_enforcement_reads_back`] fails at the
|
||||
//! transition wait or the plaintext round-trip.
|
||||
//!
|
||||
//! The replication half of the same acceptance item lives in
|
||||
//! `crates/e2e_test/src/replication_extension_test.rs`
|
||||
//! (`test_bucket_replication_sse_kms_failure_contract`); ILM had no coverage
|
||||
//! before this file.
|
||||
//!
|
||||
//! Deployment constraint pinned by the transition test's setup: the RustFS warm
|
||||
//! backend forwards the object's stored `x-amz-server-side-encryption*` metadata
|
||||
//! as raw headers on the tier data PUT (`build_transition_put_options` +
|
||||
//! `api_put_object.rs` header mapping), so a RustFS tier target must itself have
|
||||
//! KMS enabled and hold the named key or it rejects every transition upload with
|
||||
//! 400 InvalidRequest. That rejection is independent of the enforcement switch;
|
||||
//! the cold server here therefore runs its own Local KMS with the same key id.
|
||||
|
||||
use super::common::{LocalKMSTestEnvironment, create_key_with_specific_id};
|
||||
use crate::common::{RustFSTestEnvironment, admin_request, init_logging};
|
||||
use aws_sdk_s3::Client;
|
||||
use aws_sdk_s3::primitives::ByteStream;
|
||||
use aws_sdk_s3::types::{
|
||||
BucketLifecycleConfiguration, ExpirationStatus, LifecycleExpiration, LifecycleRule, LifecycleRuleFilter, RestoreRequest,
|
||||
ServerSideEncryption, ServerSideEncryptionByDefault, ServerSideEncryptionConfiguration, ServerSideEncryptionRule, Transition,
|
||||
TransitionStorageClass,
|
||||
};
|
||||
use serde::Deserialize;
|
||||
use serial_test::serial;
|
||||
use std::time::{Duration as StdDuration, Instant};
|
||||
use tracing::info;
|
||||
|
||||
type TestResult = Result<(), Box<dyn std::error::Error + Send + Sync>>;
|
||||
|
||||
const SSE_KEY: &str = "kms-ilm-sse-key";
|
||||
const PAYLOAD: &[u8] = b"kms ilm sse payload: survives enforcement, expires and transitions on schedule";
|
||||
|
||||
const EXPIRY_BUCKET: &str = "kms-ilm-expiry";
|
||||
const EXPIRE_KEY: &str = "expire/object.bin";
|
||||
const SURVIVOR_KEY: &str = "keep/object.bin";
|
||||
|
||||
const TIER_NAME: &str = "KMSCOLD";
|
||||
const TIER_BUCKET: &str = "kms-ilm-cold-tier";
|
||||
const TIER_PREFIX: &str = "tiered";
|
||||
const TRANSITION_BUCKET: &str = "kms-ilm-transition";
|
||||
const TRANSITION_KEY: &str = "tier/object.bin";
|
||||
|
||||
/// Generous CI safety net; with a 1s scanner cycle and 2s lifecycle days the
|
||||
/// terminal state normally lands within a few seconds.
|
||||
const ILM_DEADLINE: StdDuration = StdDuration::from_secs(90);
|
||||
|
||||
/// Start a Local-KMS server with per-key SSE authorization enforced and the
|
||||
/// lifecycle clock accelerated.
|
||||
///
|
||||
/// KMS wiring matches `kms_authorization_negative_matrix_test.rs` (local backend,
|
||||
/// `--kms-default-key-id`, insecure dev defaults). The lifecycle env matches
|
||||
/// `reliant/lifecycle.rs::fast_lifecycle_env` plus `RUSTFS_ILM_DEBUG_DAY_SECS=2`,
|
||||
/// so a `Days=1` rule is due about two seconds after the write.
|
||||
async fn start_enforcing_ilm_server(env: &mut LocalKMSTestEnvironment) -> TestResult {
|
||||
create_key_with_specific_id(&env.kms_keys_dir, SSE_KEY).await?;
|
||||
|
||||
let key_dir = env.kms_keys_dir.clone();
|
||||
let args = vec![
|
||||
"--kms-enable",
|
||||
"--kms-backend",
|
||||
"local",
|
||||
"--kms-key-dir",
|
||||
key_dir.as_str(),
|
||||
"--kms-default-key-id",
|
||||
SSE_KEY,
|
||||
];
|
||||
|
||||
let envs = [
|
||||
("RUSTFS_KMS_ALLOW_INSECURE_DEV_DEFAULTS", "true"),
|
||||
("RUSTFS_KMS_ENFORCE_SSE_KEY_POLICY", "true"),
|
||||
("RUSTFS_SCANNER_CYCLE", "1"),
|
||||
("RUSTFS_ILM_PROCESS_TIME", "1"),
|
||||
("RUSTFS_ILM_DEBUG_DAY_SECS", "2"),
|
||||
];
|
||||
|
||||
env.base_env.start_rustfs_server_with_env(args, &envs).await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Set the bucket's default encryption to SSE-KMS under [`SSE_KEY`], so plain
|
||||
/// PUTs (and internal rewrites) are encrypted without per-request SSE headers.
|
||||
async fn set_bucket_default_sse_kms(client: &Client, bucket: &str) -> TestResult {
|
||||
let encryption_config = ServerSideEncryptionConfiguration::builder()
|
||||
.rules(
|
||||
ServerSideEncryptionRule::builder()
|
||||
.apply_server_side_encryption_by_default(
|
||||
ServerSideEncryptionByDefault::builder()
|
||||
.sse_algorithm(ServerSideEncryption::AwsKms)
|
||||
.kms_master_key_id(SSE_KEY)
|
||||
.build()?,
|
||||
)
|
||||
.build(),
|
||||
)
|
||||
.build()?;
|
||||
client
|
||||
.put_bucket_encryption()
|
||||
.bucket(bucket)
|
||||
.server_side_encryption_configuration(encryption_config)
|
||||
.send()
|
||||
.await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Assert via `HeadObject` that the stored object is SSE-KMS encrypted under
|
||||
/// [`SSE_KEY`]. Without this, a bucket-default misconfiguration would let the
|
||||
/// tests pass on an unencrypted object and prove nothing about KMS.
|
||||
async fn assert_head_sse_kms(client: &Client, bucket: &str, key: &str) -> TestResult {
|
||||
let head = client.head_object().bucket(bucket).key(key).send().await?;
|
||||
assert_eq!(
|
||||
head.server_side_encryption(),
|
||||
Some(&ServerSideEncryption::AwsKms),
|
||||
"{bucket}/{key} must be SSE-KMS encrypted via the bucket default"
|
||||
);
|
||||
assert_eq!(
|
||||
head.ssekms_key_id(),
|
||||
Some(SSE_KEY),
|
||||
"{bucket}/{key} must be wrapped under the configured KMS key"
|
||||
);
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Returns `true` once `GET bucket/key` fails with `NoSuchKey`, `false` while it
|
||||
/// still succeeds. Any other error is surfaced. (Copied from
|
||||
/// `reliant/lifecycle.rs`; that helper is private to the reliant module.)
|
||||
async fn object_is_gone(client: &Client, bucket: &str, key: &str) -> Result<bool, Box<dyn std::error::Error + Send + Sync>> {
|
||||
match client.get_object().bucket(bucket).key(key).send().await {
|
||||
Ok(output) => {
|
||||
output.body.collect().await?;
|
||||
Ok(false)
|
||||
}
|
||||
Err(e) => {
|
||||
if let Some(service_error) = e.as_service_error() {
|
||||
if service_error.is_no_such_key() {
|
||||
return Ok(true);
|
||||
}
|
||||
return Err(format!("expected NoSuchKey, got: {e:?}").into());
|
||||
}
|
||||
Err(format!("expected a service error, got: {e:?}").into())
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/// Poll until `GET bucket/key` returns `NoSuchKey`, or fail after `deadline`.
|
||||
async fn wait_for_object_expired(client: &Client, bucket: &str, key: &str, deadline: StdDuration) -> TestResult {
|
||||
let start = Instant::now();
|
||||
loop {
|
||||
if object_is_gone(client, bucket, key).await? {
|
||||
return Ok(());
|
||||
}
|
||||
if start.elapsed() >= deadline {
|
||||
return Err(format!(
|
||||
"object {bucket}/{key} was not expired by the lifecycle scanner within {}s; \
|
||||
SSE key-policy enforcement may have started blocking the scanner's internal deletes",
|
||||
deadline.as_secs()
|
||||
)
|
||||
.into());
|
||||
}
|
||||
tokio::time::sleep(StdDuration::from_millis(500)).await;
|
||||
}
|
||||
}
|
||||
|
||||
/// Install a prefix-scoped `Days`-based expiration rule.
|
||||
async fn put_expiration_rule(client: &Client, bucket: &str, id: &str, prefix: &str, days: i32) -> TestResult {
|
||||
let rule = LifecycleRule::builder()
|
||||
.id(id)
|
||||
.filter(LifecycleRuleFilter::builder().prefix(prefix).build())
|
||||
.expiration(LifecycleExpiration::builder().days(days).build())
|
||||
.status(ExpirationStatus::Enabled)
|
||||
.build()?;
|
||||
let lifecycle = BucketLifecycleConfiguration::builder().rules(rule).build()?;
|
||||
client
|
||||
.put_bucket_lifecycle_configuration()
|
||||
.bucket(bucket)
|
||||
.lifecycle_configuration(lifecycle)
|
||||
.send()
|
||||
.await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Install a prefix-scoped `Days`-based transition rule targeting [`TIER_NAME`].
|
||||
async fn put_transition_rule(client: &Client, bucket: &str, id: &str, prefix: &str, days: i32) -> TestResult {
|
||||
let rule = LifecycleRule::builder()
|
||||
.id(id)
|
||||
.filter(LifecycleRuleFilter::builder().prefix(prefix).build())
|
||||
.transitions(
|
||||
Transition::builder()
|
||||
.days(days)
|
||||
.storage_class(TransitionStorageClass::from(TIER_NAME))
|
||||
.build(),
|
||||
)
|
||||
.status(ExpirationStatus::Enabled)
|
||||
.build()?;
|
||||
let lifecycle = BucketLifecycleConfiguration::builder().rules(rule).build()?;
|
||||
client
|
||||
.put_bucket_lifecycle_configuration()
|
||||
.bucket(bucket)
|
||||
.lifecycle_configuration(lifecycle)
|
||||
.send()
|
||||
.await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Start a plain Local-KMS server (no enforcement, no lifecycle acceleration)
|
||||
/// holding [`SSE_KEY`], to serve as the cold tier target.
|
||||
///
|
||||
/// The RustFS warm backend forwards the stored SSE-KMS headers on the tier data
|
||||
/// PUT, so the target re-applies managed SSE-KMS under the named key and must
|
||||
/// be able to resolve it; without KMS it answers 400 InvalidRequest and the
|
||||
/// transition can never complete. Enforcement stays off here: the tier writes
|
||||
/// arrive under `cold`'s root credentials, and one enforcing side is enough to
|
||||
/// pin the exemption.
|
||||
async fn start_cold_tier_kms_server(env: &mut LocalKMSTestEnvironment) -> TestResult {
|
||||
create_key_with_specific_id(&env.kms_keys_dir, SSE_KEY).await?;
|
||||
|
||||
let key_dir = env.kms_keys_dir.clone();
|
||||
let args = vec![
|
||||
"--kms-enable",
|
||||
"--kms-backend",
|
||||
"local",
|
||||
"--kms-key-dir",
|
||||
key_dir.as_str(),
|
||||
"--kms-default-key-id",
|
||||
SSE_KEY,
|
||||
];
|
||||
|
||||
env.base_env
|
||||
.start_rustfs_server_with_env(args, &[("RUSTFS_KMS_ALLOW_INSECURE_DEV_DEFAULTS", "true")])
|
||||
.await?;
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// The subset of the manual transition run report these tests assert on.
|
||||
///
|
||||
/// Unknown fields are ignored, so this stays compatible with report growth; the
|
||||
/// full shape is pinned by `reliant/tiering.rs`.
|
||||
#[derive(Debug, Deserialize)]
|
||||
struct ManualTransitionRunReport {
|
||||
#[serde(default)]
|
||||
scanned: u64,
|
||||
#[serde(default)]
|
||||
enqueued: u64,
|
||||
#[serde(default)]
|
||||
skipped_already_in_flight: u64,
|
||||
#[serde(default)]
|
||||
skipped_tier: u64,
|
||||
}
|
||||
|
||||
#[derive(Debug, Deserialize)]
|
||||
struct ManualTransitionRunResponse {
|
||||
state: String,
|
||||
report: ManualTransitionRunReport,
|
||||
}
|
||||
|
||||
/// One synchronous (enqueue-only) manual transition run over `bucket/prefix`,
|
||||
/// via the same admin endpoint `reliant/tiering.rs` drives.
|
||||
async fn manual_transition_run(
|
||||
hot: &RustFSTestEnvironment,
|
||||
bucket: &str,
|
||||
prefix: &str,
|
||||
) -> Result<ManualTransitionRunResponse, Box<dyn std::error::Error + Send + Sync>> {
|
||||
let bucket = urlencoding::encode(bucket);
|
||||
let prefix = urlencoding::encode(prefix);
|
||||
let tier = urlencoding::encode(TIER_NAME);
|
||||
let path =
|
||||
format!("/rustfs/admin/v3/ilm/transition/run?bucket={bucket}&prefix={prefix}&tier={tier}&dryRun=false&maxObjects=10");
|
||||
let (status, body) = admin_request(&hot.url, http::Method::POST, &path, None, &hot.access_key, &hot.secret_key).await?;
|
||||
if !status.is_success() {
|
||||
return Err(format!("manual transition run failed: status={status}, body={body}").into());
|
||||
}
|
||||
Ok(serde_json::from_str(&body)?)
|
||||
}
|
||||
|
||||
/// Drive manual transition runs until one reports the object as processed.
|
||||
///
|
||||
/// The `Days=1` rule becomes due about two seconds after the write
|
||||
/// (`RUSTFS_ILM_DEBUG_DAY_SECS=2`), so early runs may legitimately report the
|
||||
/// object as not yet eligible; the loop keeps running the endpoint until it
|
||||
/// either enqueues the transition, sees it already in flight (the 1s scanner
|
||||
/// backstop got there first), or finds it already on the tier.
|
||||
async fn run_manual_transition_until_processed(
|
||||
hot: &RustFSTestEnvironment,
|
||||
bucket: &str,
|
||||
prefix: &str,
|
||||
deadline: StdDuration,
|
||||
) -> TestResult {
|
||||
let start = Instant::now();
|
||||
loop {
|
||||
let run = manual_transition_run(hot, bucket, prefix).await?;
|
||||
assert_eq!(run.report.scanned, 1, "manual transition run must scan the object: {run:#?}");
|
||||
if run.report.enqueued + run.report.skipped_already_in_flight + run.report.skipped_tier >= 1 {
|
||||
info!(state = %run.state, report = ?run.report, "manual transition run processed the SSE-KMS object");
|
||||
return Ok(());
|
||||
}
|
||||
if start.elapsed() >= deadline {
|
||||
return Err(format!(
|
||||
"manual transition runs never processed {bucket}/{prefix} within {}s; last report: {run:#?}",
|
||||
deadline.as_secs()
|
||||
)
|
||||
.into());
|
||||
}
|
||||
tokio::time::sleep(StdDuration::from_millis(500)).await;
|
||||
}
|
||||
}
|
||||
|
||||
/// Wire `hot` -> `cold` as a `TierType::RustFS` remote tier via `AddTier`.
|
||||
///
|
||||
/// No `force`, so the server runs the real connectivity probe against `cold`
|
||||
/// (the tier bucket must already exist there). Mirrors
|
||||
/// `reliant/tiering.rs::add_rustfs_tier`, which is private to that module.
|
||||
async fn add_rustfs_tier(hot: &RustFSTestEnvironment, cold: &RustFSTestEnvironment) -> TestResult {
|
||||
let body = serde_json::json!({
|
||||
"type": "rustfs",
|
||||
"rustfs": {
|
||||
"name": TIER_NAME,
|
||||
"endpoint": cold.url.as_str(),
|
||||
"accessKey": cold.access_key.as_str(),
|
||||
"secretKey": cold.secret_key.as_str(),
|
||||
"bucket": TIER_BUCKET,
|
||||
"prefix": TIER_PREFIX,
|
||||
"region": "us-east-1",
|
||||
"storageClass": ""
|
||||
}
|
||||
})
|
||||
.to_string();
|
||||
|
||||
let (status, resp) = admin_request(
|
||||
&hot.url,
|
||||
http::Method::PUT,
|
||||
"/rustfs/admin/v3/tier",
|
||||
Some(body),
|
||||
&hot.access_key,
|
||||
&hot.secret_key,
|
||||
)
|
||||
.await?;
|
||||
if !status.is_success() {
|
||||
return Err(format!("AddTier(RustFS) failed: status={status}, body={resp}").into());
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Poll `HEAD` until the object's storage class is the tier name (transition
|
||||
/// complete), or fail after `deadline`. (From `reliant/tiering.rs`.)
|
||||
async fn wait_for_transition(client: &Client, bucket: &str, key: &str, deadline: StdDuration) -> TestResult {
|
||||
let start = Instant::now();
|
||||
loop {
|
||||
let head = client.head_object().bucket(bucket).key(key).send().await?;
|
||||
if head.storage_class().map(|sc| sc.as_str()) == Some(TIER_NAME) {
|
||||
return Ok(());
|
||||
}
|
||||
if start.elapsed() >= deadline {
|
||||
return Err(format!(
|
||||
"object {bucket}/{key} was not transitioned to {TIER_NAME} within {}s (storage_class={:?}); \
|
||||
SSE key-policy enforcement may have started blocking the transition worker's internal reads",
|
||||
deadline.as_secs(),
|
||||
head.storage_class()
|
||||
)
|
||||
.into());
|
||||
}
|
||||
tokio::time::sleep(StdDuration::from_millis(500)).await;
|
||||
}
|
||||
}
|
||||
|
||||
/// Poll `HEAD` until `x-amz-restore` reports a finished restore
|
||||
/// (`ongoing-request="false"`), or fail after `deadline`.
|
||||
async fn wait_for_restore_complete(client: &Client, bucket: &str, key: &str, deadline: StdDuration) -> TestResult {
|
||||
let start = Instant::now();
|
||||
loop {
|
||||
let head = client.head_object().bucket(bucket).key(key).send().await?;
|
||||
if head.restore().is_some_and(|r| r.contains("ongoing-request=\"false\"")) {
|
||||
return Ok(());
|
||||
}
|
||||
if start.elapsed() >= deadline {
|
||||
return Err(format!(
|
||||
"object {bucket}/{key} restore did not complete within {}s (restore={:?}); \
|
||||
SSE key-policy enforcement may have started blocking the restore copy-back's internal reads",
|
||||
deadline.as_secs(),
|
||||
head.restore()
|
||||
)
|
||||
.into());
|
||||
}
|
||||
tokio::time::sleep(StdDuration::from_millis(500)).await;
|
||||
}
|
||||
}
|
||||
|
||||
/// ILM expiration keeps working on an SSE-KMS bucket while per-key SSE
|
||||
/// authorization is enforced.
|
||||
///
|
||||
/// The lifecycle scanner deletes expired objects with an internal (no-principal)
|
||||
/// identity that holds no `kms` grant. If enforcement ever starts applying to
|
||||
/// those internal deletes (or to the scanner's metadata reads) on encrypted
|
||||
/// buckets, expiry stops happening and this test times out.
|
||||
///
|
||||
/// A survivor object under a non-matching prefix isolates the rule's prefix
|
||||
/// filter as the cause of the deletion and proves the encrypted bucket stays
|
||||
/// readable end to end after the scanner has run.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn ilm_expiration_on_sse_kms_bucket_under_enforcement() -> TestResult {
|
||||
init_logging();
|
||||
|
||||
let mut env = LocalKMSTestEnvironment::new().await?;
|
||||
start_enforcing_ilm_server(&mut env).await?;
|
||||
env.base_env.create_test_bucket(EXPIRY_BUCKET).await?;
|
||||
|
||||
let client = env.base_env.create_s3_client();
|
||||
set_bucket_default_sse_kms(&client, EXPIRY_BUCKET).await?;
|
||||
|
||||
for key in [EXPIRE_KEY, SURVIVOR_KEY] {
|
||||
client
|
||||
.put_object()
|
||||
.bucket(EXPIRY_BUCKET)
|
||||
.key(key)
|
||||
.body(ByteStream::from_static(PAYLOAD))
|
||||
.send()
|
||||
.await?;
|
||||
assert_head_sse_kms(&client, EXPIRY_BUCKET, key).await?;
|
||||
}
|
||||
info!("both objects stored SSE-KMS encrypted under enforcement");
|
||||
|
||||
put_expiration_rule(&client, EXPIRY_BUCKET, "kms-ilm-expire", "expire/", 1).await?;
|
||||
|
||||
// The regression this pins: the scanner's internal delete must stay exempt
|
||||
// from per-key SSE authorization, so the encrypted object actually expires.
|
||||
wait_for_object_expired(&client, EXPIRY_BUCKET, EXPIRE_KEY, ILM_DEADLINE).await?;
|
||||
info!("SSE-KMS object expired by the lifecycle scanner under enforcement");
|
||||
|
||||
// Negative control: same bucket, same encryption, non-matching prefix. It
|
||||
// must survive the scanner and still decrypt for the requesting principal.
|
||||
assert!(
|
||||
!object_is_gone(&client, EXPIRY_BUCKET, SURVIVOR_KEY).await?,
|
||||
"non-matching-prefix object must not be expired by a prefix-scoped rule"
|
||||
);
|
||||
let survivor = client.get_object().bucket(EXPIRY_BUCKET).key(SURVIVOR_KEY).send().await?;
|
||||
assert_eq!(
|
||||
survivor.body.collect().await?.into_bytes().as_ref(),
|
||||
PAYLOAD,
|
||||
"surviving SSE-KMS object must still decrypt after the scanner has run"
|
||||
);
|
||||
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// ILM transition to a remote tier keeps working on an SSE-KMS bucket while
|
||||
/// per-key SSE authorization is enforced, and the transitioned object reads
|
||||
/// back as plaintext.
|
||||
///
|
||||
/// The transition worker moves the stored (encrypted) bytes to the cold tier
|
||||
/// with an internal (no-principal) identity; the read-through `GET` then
|
||||
/// decrypts the envelope for the requesting principal. If enforcement ever
|
||||
/// starts applying to the worker's internal reads, the transition wait times
|
||||
/// out; if the stored envelope is mishandled across the tier round trip, the
|
||||
/// plaintext comparison fails.
|
||||
///
|
||||
/// The transition is driven through the manual transition-run admin endpoint
|
||||
/// (the mechanism `reliant/tiering.rs` established), so the test does not
|
||||
/// depend on scanner scheduling; the 1s scanner cycle stays on as a backstop.
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn ilm_transition_on_sse_kms_bucket_under_enforcement_reads_back() -> TestResult {
|
||||
init_logging();
|
||||
|
||||
// Cold-tier server: independent credentials, its own Local KMS holding the
|
||||
// same key id (see the module docs for why the tier target needs KMS).
|
||||
// Started first; each server's startup cleanup only matches its own unique
|
||||
// address and temp dir, so the two instances coexist.
|
||||
let mut cold = LocalKMSTestEnvironment::new().await?;
|
||||
cold.base_env.access_key = "kmscoldtieradmin".to_string();
|
||||
cold.base_env.secret_key = "kmscoldtiersecret".to_string();
|
||||
start_cold_tier_kms_server(&mut cold).await?;
|
||||
let cold_client = cold.base_env.create_s3_client();
|
||||
cold_client.create_bucket().bucket(TIER_BUCKET).send().await?;
|
||||
|
||||
// Hot server: Local KMS + enforcement + accelerated lifecycle clock.
|
||||
let mut env = LocalKMSTestEnvironment::new().await?;
|
||||
start_enforcing_ilm_server(&mut env).await?;
|
||||
let hot_client = env.base_env.create_s3_client();
|
||||
|
||||
add_rustfs_tier(&env.base_env, &cold.base_env).await?;
|
||||
|
||||
env.base_env.create_test_bucket(TRANSITION_BUCKET).await?;
|
||||
set_bucket_default_sse_kms(&hot_client, TRANSITION_BUCKET).await?;
|
||||
|
||||
hot_client
|
||||
.put_object()
|
||||
.bucket(TRANSITION_BUCKET)
|
||||
.key(TRANSITION_KEY)
|
||||
.body(ByteStream::from_static(PAYLOAD))
|
||||
.send()
|
||||
.await?;
|
||||
assert_head_sse_kms(&hot_client, TRANSITION_BUCKET, TRANSITION_KEY).await?;
|
||||
info!("object stored SSE-KMS encrypted under enforcement");
|
||||
|
||||
// Days=1 is due ~2s after the write with RUSTFS_ILM_DEBUG_DAY_SECS=2.
|
||||
put_transition_rule(&hot_client, TRANSITION_BUCKET, "kms-ilm-transition", "tier/", 1).await?;
|
||||
|
||||
// Drive the transition deterministically via the manual run endpoint, then
|
||||
// wait for HEAD to report the tier as the object's storage class.
|
||||
run_manual_transition_until_processed(&env.base_env, TRANSITION_BUCKET, "tier/", ILM_DEADLINE).await?;
|
||||
wait_for_transition(&hot_client, TRANSITION_BUCKET, TRANSITION_KEY, ILM_DEADLINE).await?;
|
||||
info!("SSE-KMS object transitioned to the remote tier under enforcement");
|
||||
|
||||
let head = hot_client
|
||||
.head_object()
|
||||
.bucket(TRANSITION_BUCKET)
|
||||
.key(TRANSITION_KEY)
|
||||
.send()
|
||||
.await?;
|
||||
assert!(
|
||||
head.restore().is_none(),
|
||||
"a freshly transitioned object must not advertise x-amz-restore, got {:?}",
|
||||
head.restore()
|
||||
);
|
||||
|
||||
// The remote copy exists on the cold tier. The payload the tier holds is the
|
||||
// hot server's stored ciphertext, wrapped once more under the cold server's
|
||||
// own managed SSE-KMS layer (the forwarded headers re-request encryption).
|
||||
let remote = cold_client.list_objects_v2().bucket(TIER_BUCKET).send().await?;
|
||||
assert!(!remote.contents().is_empty(), "cold-tier bucket must hold the transitioned object's data");
|
||||
|
||||
// Read-through GET under enforcement must succeed (not AccessDenied) and
|
||||
// keep advertising SSE-KMS. Its BODY is deliberately not compared here:
|
||||
// the transitioned read path skips managed-SSE decryption — a product gap
|
||||
// unrelated to enforcement — so a direct GET streams the stored ciphertext
|
||||
// (`new_getobjectreader` in crates/ecstore/src/client/object_api_utils.rs
|
||||
// hardcodes `is_encrypted = false` and never applies the
|
||||
// `ReadTransform::Encrypted` wrapping the hot-read path builds in
|
||||
// crates/ecstore/src/object_api/readers.rs). Plaintext recovery is pinned
|
||||
// through restore semantics below; when the read-through gap is fixed, a
|
||||
// byte assertion can be added here too.
|
||||
let read_through = hot_client
|
||||
.get_object()
|
||||
.bucket(TRANSITION_BUCKET)
|
||||
.key(TRANSITION_KEY)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
read_through.server_side_encryption(),
|
||||
Some(&ServerSideEncryption::AwsKms),
|
||||
"transitioned object must still report SSE-KMS on read-through"
|
||||
);
|
||||
let read_through_body = read_through.body.collect().await?.into_bytes();
|
||||
assert_eq!(
|
||||
read_through_body.len(),
|
||||
PAYLOAD.len(),
|
||||
"read-through GET must stream the object's full logical size under enforcement"
|
||||
);
|
||||
|
||||
// RestoreObject copies the ciphertext back from the tier under the original
|
||||
// envelope metadata; the restored copy is then served by the normal
|
||||
// decrypting read path. The copy-back runs with an internal (no-principal)
|
||||
// identity, so this also pins the exemption on the restore path. Days=300
|
||||
// because RUSTFS_ILM_DEBUG_DAY_SECS=2 accelerates the restored copy's
|
||||
// expiry as well (300 accelerated days == 600s of validity).
|
||||
hot_client
|
||||
.restore_object()
|
||||
.bucket(TRANSITION_BUCKET)
|
||||
.key(TRANSITION_KEY)
|
||||
.restore_request(RestoreRequest::builder().days(300).build())
|
||||
.send()
|
||||
.await?;
|
||||
wait_for_restore_complete(&hot_client, TRANSITION_BUCKET, TRANSITION_KEY, ILM_DEADLINE).await?;
|
||||
info!("SSE-KMS object restored from the remote tier under enforcement");
|
||||
|
||||
// The KMS-relevant half: the restored envelope decrypts back to the exact
|
||||
// plaintext for the requesting principal.
|
||||
let restored = hot_client
|
||||
.get_object()
|
||||
.bucket(TRANSITION_BUCKET)
|
||||
.key(TRANSITION_KEY)
|
||||
.send()
|
||||
.await?;
|
||||
assert_eq!(
|
||||
restored.server_side_encryption(),
|
||||
Some(&ServerSideEncryption::AwsKms),
|
||||
"restored object must still report SSE-KMS"
|
||||
);
|
||||
let body = restored.body.collect().await?.into_bytes();
|
||||
assert_eq!(body.as_ref(), PAYLOAD, "restored SSE-KMS object must round-trip byte-identical plaintext");
|
||||
|
||||
Ok(())
|
||||
}
|
||||
@@ -59,6 +59,3 @@ mod configured_roundtrip_test;
|
||||
|
||||
#[cfg(test)]
|
||||
mod kms_authorization_negative_matrix_test;
|
||||
|
||||
#[cfg(test)]
|
||||
mod kms_ilm_sse_kms_test;
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -2854,7 +2854,7 @@ pub(crate) mod cmptst_30 {
|
||||
result
|
||||
}
|
||||
|
||||
#[ignore = "timing-sensitive backend-pressure latency probe; run explicitly with --ignored"]
|
||||
#[ignore]
|
||||
#[tokio::test]
|
||||
async fn regression() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
||||
crate::common::init_logging();
|
||||
|
||||
@@ -2401,20 +2401,15 @@ async fn wait_for_site_replication_info<F>(
|
||||
where
|
||||
F: Fn(&SiteReplicationInfo) -> bool,
|
||||
{
|
||||
// 30s to match wait_for_replication_state: the three-node site tests run
|
||||
// several full rustfs processes on one runner, so peer-state propagation
|
||||
// can take well over 10s under CI load.
|
||||
let deadline = tokio::time::Instant::now() + Duration::from_secs(30);
|
||||
loop {
|
||||
for _ in 0..40 {
|
||||
let info = site_replication_info(env).await?;
|
||||
if predicate(&info) {
|
||||
return Ok(info);
|
||||
}
|
||||
if tokio::time::Instant::now() >= deadline {
|
||||
return Err(format!("site replication info did not reach expected state on {}", env.address).into());
|
||||
}
|
||||
sleep(Duration::from_millis(250)).await;
|
||||
}
|
||||
|
||||
Err(format!("site replication info did not reach expected state on {}", env.address).into())
|
||||
}
|
||||
|
||||
async fn wait_for_site_replication_status<F>(
|
||||
@@ -2425,19 +2420,15 @@ async fn wait_for_site_replication_status<F>(
|
||||
where
|
||||
F: Fn(&SRStatusInfo) -> bool,
|
||||
{
|
||||
// Same 30s ceiling as wait_for_site_replication_info: the status probes
|
||||
// fan out to every peer, so they see the same multi-process CI load.
|
||||
let deadline = tokio::time::Instant::now() + Duration::from_secs(30);
|
||||
loop {
|
||||
for _ in 0..40 {
|
||||
let status = site_replication_status(env, query).await?;
|
||||
if predicate(&status) {
|
||||
return Ok(status);
|
||||
}
|
||||
if tokio::time::Instant::now() >= deadline {
|
||||
return Err(format!("site replication status did not reach expected state on {}", env.address).into());
|
||||
}
|
||||
sleep(Duration::from_millis(250)).await;
|
||||
}
|
||||
|
||||
Err(format!("site replication status did not reach expected state on {}", env.address).into())
|
||||
}
|
||||
|
||||
async fn wait_for_replication_reset_target<F>(
|
||||
@@ -4244,49 +4235,37 @@ async fn test_bucket_replication_acceptance_matrix_local_dual_targets() -> TestR
|
||||
"tag rule with disabled delete-marker replication created a marker: {tagged_state:?}"
|
||||
);
|
||||
|
||||
// AWS S3 and MinIO both reject suspending versioning on a bucket that
|
||||
// carries a replication configuration (InvalidBucketState): suspension
|
||||
// would mint null versions that versioned replication can never converge.
|
||||
let suspend_err = source_client
|
||||
.put_bucket_versioning()
|
||||
.bucket(source_bucket)
|
||||
.versioning_configuration(
|
||||
VersioningConfiguration::builder()
|
||||
.status(BucketVersioningStatus::Suspended)
|
||||
.build(),
|
||||
)
|
||||
.send()
|
||||
.await
|
||||
.expect_err("suspending versioning on a replication source must be rejected");
|
||||
assert_eq!(
|
||||
suspend_err.as_service_error().and_then(|error| error.code()),
|
||||
Some("InvalidBucketState"),
|
||||
"suspension on a replication source must fail with InvalidBucketState: {suspend_err:?}"
|
||||
);
|
||||
|
||||
// The rejected suspension must leave the versioning + replication state
|
||||
// fully intact: a fresh matched PUT still replicates with a real version.
|
||||
let post_reject_put = source_client
|
||||
set_bucket_versioning(&source_env, source_bucket, BucketVersioningStatus::Suspended).await?;
|
||||
set_bucket_versioning(&target_env_a, target_bucket_a, BucketVersioningStatus::Suspended).await?;
|
||||
let null_put = source_client
|
||||
.put_object()
|
||||
.bucket(source_bucket)
|
||||
.key("prefix/after-rejected-suspend.txt")
|
||||
.body(ByteStream::from_static(b"still replicating"))
|
||||
.key("prefix/null.txt")
|
||||
.body(ByteStream::from_static(b"null version"))
|
||||
.send()
|
||||
.await?;
|
||||
let post_reject_version_id = post_reject_put
|
||||
.version_id()
|
||||
.ok_or("PUT after rejected suspension omitted version ID")?
|
||||
.to_string();
|
||||
wait_for_replication_state(
|
||||
&target_client_a,
|
||||
target_bucket_a,
|
||||
"replication stopped after rejected versioning suspension",
|
||||
|state| {
|
||||
state
|
||||
.iter()
|
||||
.any(|entry| entry.key == "prefix/after-rejected-suspend.txt" && entry.version_id == post_reject_version_id)
|
||||
},
|
||||
)
|
||||
assert!(null_put.version_id().is_none(), "suspended source PUT must create a null version");
|
||||
wait_for_replication_state(&target_client_a, target_bucket_a, "null version did not replicate", |state| {
|
||||
state
|
||||
.iter()
|
||||
.any(|entry| entry.key == "prefix/null.txt" && entry.version_id == "null" && !entry.delete_marker)
|
||||
})
|
||||
.await?;
|
||||
let null_delete = source_client
|
||||
.delete_object()
|
||||
.bucket(source_bucket)
|
||||
.key("prefix/null.txt")
|
||||
.send()
|
||||
.await?;
|
||||
assert!(
|
||||
null_delete.version_id().is_none(),
|
||||
"suspended source DELETE must create a null delete marker"
|
||||
);
|
||||
wait_for_replication_state(&target_client_a, target_bucket_a, "null delete marker did not replicate", |state| {
|
||||
state
|
||||
.iter()
|
||||
.any(|entry| entry.key == "prefix/null.txt" && entry.version_id == "null" && entry.delete_marker)
|
||||
})
|
||||
.await?;
|
||||
|
||||
Ok(())
|
||||
|
||||
@@ -32,11 +32,6 @@ workspace = true
|
||||
|
||||
[features]
|
||||
default = []
|
||||
# Compiles the controlled list-objects namespace-journal chaos injector into a
|
||||
# production binary (it is always available to tests). Off by default so the
|
||||
# RUSTFS_LIST_OBJECTS_NAMESPACE_JOURNAL_CHAOS_* env vars cannot rewrite journal
|
||||
# state in a stock build (backlog#1832).
|
||||
list-chaos = []
|
||||
rio-v2 = ["dep:rustfs-rio-v2"]
|
||||
hotpath = [
|
||||
"hotpath/hotpath",
|
||||
|
||||
@@ -61,11 +61,9 @@ pub mod bucket {
|
||||
delete_manual_transition_scope_admission_if_current, load_manual_transition_job_record,
|
||||
load_manual_transition_job_record_with_etag, load_manual_transition_scope_admission,
|
||||
manual_transition_job_lease_expired, manual_transition_scope_admission_lease_expired,
|
||||
manual_transition_scope_key, persist_manual_transition_job_progress,
|
||||
persist_manual_transition_job_progress_if_owned, renew_manual_transition_job_lease,
|
||||
renew_manual_transition_job_lease_if_owned, request_manual_transition_job_cancel,
|
||||
save_manual_transition_job_record, save_manual_transition_job_record_if_current,
|
||||
save_manual_transition_scope_admission_if_absent, update_manual_transition_job_record,
|
||||
manual_transition_scope_key, persist_manual_transition_job_progress, renew_manual_transition_job_lease,
|
||||
request_manual_transition_job_cancel, save_manual_transition_job_record,
|
||||
save_manual_transition_job_record_if_current, save_manual_transition_scope_admission_if_absent,
|
||||
};
|
||||
}
|
||||
|
||||
@@ -131,19 +129,16 @@ pub mod bucket {
|
||||
}
|
||||
|
||||
pub mod metadata_sys {
|
||||
#[cfg(feature = "test-util")]
|
||||
pub use crate::bucket::metadata_sys::ConfigWriteLockProbe;
|
||||
pub use crate::bucket::metadata_sys::{
|
||||
BucketMetadataMutationGuard, BucketMetadataSys, ObjectLockConfigState, acquire_bucket_metadata_transaction_lock,
|
||||
acquire_bucket_metadata_transaction_lock_for_incarnation, capture_bucket_metadata_incarnation, delete,
|
||||
delete_if_incarnation, delete_under_transaction_lock, get, get_accelerate_config, get_bucket_policy,
|
||||
capture_bucket_metadata_incarnation, delete, delete_if_incarnation, get, get_accelerate_config, get_bucket_policy,
|
||||
get_bucket_policy_raw, get_bucket_targets_config, get_config_from_disk, get_cors_config, get_durability_config,
|
||||
get_global_bucket_metadata_sys, get_lifecycle_config, get_logging_config, get_notification_config,
|
||||
get_object_lock_config, get_object_lock_config_state, get_public_access_block_config, get_quota_config,
|
||||
get_replication_config, get_request_payment_config, get_sse_config, get_tagging_config, get_versioning_config,
|
||||
get_website_config, init_bucket_metadata_sys, list_bucket_targets, reload_bucket_metadata, remove_bucket_metadata,
|
||||
set_bucket_metadata, update, update_bucket_targets_under_transaction_lock, update_config_with, update_if_incarnation,
|
||||
update_quota_if_incarnation, update_under_transaction_lock,
|
||||
update_under_transaction_lock,
|
||||
};
|
||||
}
|
||||
|
||||
@@ -185,18 +180,17 @@ pub mod bucket {
|
||||
mrf_backlog_observability_snapshot,
|
||||
};
|
||||
pub use crate::bucket::replication::{
|
||||
BucketReplicationResyncStatus, BucketReplicationStat, BucketReplicationStats, BucketStats,
|
||||
DeleteReplicationConfigSnapshot, DeletedObjectReplicationInfo, DurableMrfBacklog, DynReplicationPool, InQueueMetric,
|
||||
MrfOpKind, MrfReplicateEntry, MustReplicateOptions, ObjectOpts, REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION,
|
||||
REMOTE_TARGET_UNSUPPORTED_FIELDS, REMOTE_TARGET_WRITABLE_FIELDS, REPLICATE_INCOMING_DELETE,
|
||||
REPLICATION_CAPABILITY_CONTRACT_VERSION, REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS,
|
||||
ReplicateDecision, ReplicateObjectInfo, ReplicationBatchAdmission, ReplicationConfig,
|
||||
ReplicationConfigStructureError, ReplicationConfigurationExt, ReplicationDeleteScheduleInput,
|
||||
ReplicationDeleteStateSource, ReplicationHealQueueResult, ReplicationObjectBridge, ReplicationObjectIO,
|
||||
ReplicationOperation, ReplicationPoolTrait, ReplicationPriority, ReplicationQueueAdmission, ReplicationScannerBridge,
|
||||
ReplicationState, ReplicationStats, ReplicationStatusType, ReplicationStorage, ReplicationTargetValidationError,
|
||||
ReplicationType, ResyncOpts, ResyncStatusType, RuntimeReplicationTargetBacklog, TargetReplicationResyncStatus,
|
||||
VersionPurgeStatusType, XferStats, commit_force_delete_intent, complete_force_delete_intent,
|
||||
BucketReplicationResyncStatus, BucketReplicationStats, BucketStats, DeleteReplicationConfigSnapshot,
|
||||
DeletedObjectReplicationInfo, DurableMrfBacklog, DynReplicationPool, MrfOpKind, MrfReplicateEntry,
|
||||
MustReplicateOptions, ObjectOpts, REMOTE_TARGET_CAPABILITY_CONTRACT_VERSION, REMOTE_TARGET_UNSUPPORTED_FIELDS,
|
||||
REMOTE_TARGET_WRITABLE_FIELDS, REPLICATE_INCOMING_DELETE, REPLICATION_CAPABILITY_CONTRACT_VERSION,
|
||||
REPLICATION_READ_ONLY_HISTORICAL_FIELDS, REPLICATION_WRITABLE_FIELDS, ReplicateDecision, ReplicateObjectInfo,
|
||||
ReplicationBatchAdmission, ReplicationConfig, ReplicationConfigStructureError, ReplicationConfigurationExt,
|
||||
ReplicationDeleteScheduleInput, ReplicationDeleteStateSource, ReplicationHealQueueResult, ReplicationObjectBridge,
|
||||
ReplicationObjectIO, ReplicationOperation, ReplicationPoolTrait, ReplicationPriority, ReplicationQueueAdmission,
|
||||
ReplicationScannerBridge, ReplicationState, ReplicationStats, ReplicationStatusType, ReplicationStorage,
|
||||
ReplicationTargetValidationError, ReplicationType, ResyncOpts, ResyncStatusType, RuntimeReplicationTargetBacklog,
|
||||
TargetReplicationResyncStatus, VersionPurgeStatusType, commit_force_delete_intent, complete_force_delete_intent,
|
||||
delete_replication_state_from_config, delete_replication_version_id, get_global_replication_pool,
|
||||
get_global_replication_stats, init_background_replication, invalid_replication_config_status_field,
|
||||
persist_force_delete_intent, read_durable_mrf_backlog, replication_state_to_filemeta, replication_status_to_filemeta,
|
||||
@@ -280,9 +274,7 @@ pub mod cluster {
|
||||
}
|
||||
|
||||
pub mod compression {
|
||||
pub use crate::io_support::compress::{
|
||||
MIN_DISK_COMPRESSIBLE_SIZE, is_disk_compressible, is_disk_compression_enabled, is_multipart_disk_compression_enabled,
|
||||
};
|
||||
pub use crate::io_support::compress::{MIN_DISK_COMPRESSIBLE_SIZE, is_disk_compressible, is_disk_compression_enabled};
|
||||
}
|
||||
|
||||
pub mod config {
|
||||
@@ -322,7 +314,7 @@ pub mod data_usage {
|
||||
DATA_USAGE_CACHE_NAME, apply_bucket_usage_memory_overlay, compute_bucket_usage,
|
||||
init_compression_total_memory_from_backend, invalidate_admin_data_usage_snapshot_cache,
|
||||
invalidate_data_usage_snapshot_cache, live_bucket_usage_computations, load_admin_data_usage_from_backend_cached,
|
||||
load_compression_total_from_memory, load_data_usage_from_backend, load_data_usage_from_backend_cached, quota_object_size,
|
||||
load_compression_total_from_memory, load_data_usage_from_backend, load_data_usage_from_backend_cached,
|
||||
record_bucket_delete_marker_memory, record_bucket_object_delete_memory, record_bucket_object_version_write_memory,
|
||||
record_bucket_object_write_memory, record_bucket_object_write_unknown_previous_memory, record_compression_total_memory,
|
||||
refresh_bucket_usage_from_object_layer, refresh_versioned_bucket_usage_from_object_layer,
|
||||
@@ -352,7 +344,7 @@ pub mod disk {
|
||||
}
|
||||
|
||||
pub mod error {
|
||||
pub use crate::disk::error::{DiskError, Error, FileAccessDeniedWithContext, Result};
|
||||
pub use crate::disk::error::{BitrotErrorType, DiskError, Error, FileAccessDeniedWithContext, Result};
|
||||
}
|
||||
|
||||
pub mod error_reduce {
|
||||
@@ -409,11 +401,8 @@ pub mod metrics {
|
||||
}
|
||||
|
||||
pub mod notification {
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
pub use crate::services::notification_sys::rotate_cross_pool_fence_fleet_proof_for_test;
|
||||
pub use crate::services::notification_sys::{
|
||||
CrossPoolFenceFleetProofToken, NotificationPeerErr, NotificationSys, acquire_cross_pool_fence_fleet_proof,
|
||||
cross_pool_fence_fleet_proof_matches, get_global_notification_sys, new_global_notification_sys,
|
||||
NotificationPeerErr, NotificationSys, get_global_notification_sys, new_global_notification_sys,
|
||||
start_remote_version_state_fleet_probe,
|
||||
};
|
||||
}
|
||||
@@ -473,8 +462,7 @@ pub mod set_disk {
|
||||
|
||||
#[cfg(feature = "test-util")]
|
||||
pub mod test_util {
|
||||
pub use crate::bucket::quota::reservation::fail_next_quota_ledger_save_for_test;
|
||||
pub use crate::set_disk::{MultipartCommitBarrier, MultipartCommitPause, PutObjectCommitBarrier, PutObjectCommitPause};
|
||||
pub use crate::set_disk::{PutObjectCommitBarrier, PutObjectCommitPause};
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@@ -58,9 +58,7 @@ use rustfs_utils::http::{
|
||||
};
|
||||
use rustfs_utils::http::{
|
||||
SUFFIX_FORCE_DELETE, SUFFIX_SOURCE_DELETEMARKER, SUFFIX_SOURCE_ETAG, SUFFIX_SOURCE_MTIME, SUFFIX_SOURCE_REPLICATION_CHECK,
|
||||
SUFFIX_SOURCE_REPLICATION_LEGALHOLD_TIMESTAMP, SUFFIX_SOURCE_REPLICATION_REQUEST,
|
||||
SUFFIX_SOURCE_REPLICATION_RETENTION_TIMESTAMP, SUFFIX_SOURCE_REPLICATION_TAGGING_TIMESTAMP, SUFFIX_SOURCE_VERSION_ID,
|
||||
insert_header,
|
||||
SUFFIX_SOURCE_REPLICATION_REQUEST, SUFFIX_SOURCE_VERSION_ID, insert_header,
|
||||
};
|
||||
use rustls_pki_types::pem::PemObject;
|
||||
use serde::{Deserialize, Serialize};
|
||||
@@ -82,6 +80,7 @@ use tracing::warn;
|
||||
use url::Url;
|
||||
use uuid::Uuid;
|
||||
|
||||
const DEFAULT_HEALTH_CHECK_RELOAD_DURATION: Duration = Duration::from_secs(30 * 60);
|
||||
const MAX_CONCURRENT_TARGET_HEALTH_CHECKS: usize = 16;
|
||||
const REDACTED_CREDENTIAL: &str = "<redacted>";
|
||||
|
||||
@@ -1477,12 +1476,9 @@ impl Default for AdvancedPutOptions {
|
||||
replication_status: ReplicationStatusType::Pending,
|
||||
source_mtime: OffsetDateTime::now_utc(),
|
||||
replication_request: false,
|
||||
// UNIX_EPOCH means "never modified": header() must not emit a
|
||||
// timestamp header for it, otherwise a receiver would treat an
|
||||
// unset category as a modification made right now.
|
||||
retention_timestamp: OffsetDateTime::UNIX_EPOCH,
|
||||
tagging_timestamp: OffsetDateTime::UNIX_EPOCH,
|
||||
legalhold_timestamp: OffsetDateTime::UNIX_EPOCH,
|
||||
retention_timestamp: OffsetDateTime::now_utc(),
|
||||
tagging_timestamp: OffsetDateTime::now_utc(),
|
||||
legalhold_timestamp: OffsetDateTime::now_utc(),
|
||||
replication_validity_check: false,
|
||||
}
|
||||
}
|
||||
@@ -1679,16 +1675,6 @@ impl PutObjectOptions {
|
||||
);
|
||||
}
|
||||
|
||||
for (suffix, timestamp) in [
|
||||
(SUFFIX_SOURCE_REPLICATION_TAGGING_TIMESTAMP, self.internal.tagging_timestamp),
|
||||
(SUFFIX_SOURCE_REPLICATION_RETENTION_TIMESTAMP, self.internal.retention_timestamp),
|
||||
(SUFFIX_SOURCE_REPLICATION_LEGALHOLD_TIMESTAMP, self.internal.legalhold_timestamp),
|
||||
] {
|
||||
if timestamp.unix_timestamp() != 0 {
|
||||
insert_header(&mut header, suffix, timestamp.format(&Rfc3339).unwrap_or_default());
|
||||
}
|
||||
}
|
||||
|
||||
if self.internal.replication_request {
|
||||
insert_header(&mut header, SUFFIX_SOURCE_REPLICATION_REQUEST, "true");
|
||||
}
|
||||
@@ -2856,57 +2842,6 @@ mod tests {
|
||||
);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn put_object_headers_carry_replication_timestamp_headers() {
|
||||
// MinIO receivers resolve concurrent tag/retention/legal-hold edits by
|
||||
// last-writer-wins on these headers (object-api-options.go parses them
|
||||
// as RFC3339); a replica without them loses every conflict resolution.
|
||||
let mut opts = PutObjectOptions::default();
|
||||
opts.internal.replication_request = true;
|
||||
let tagging = OffsetDateTime::from_unix_timestamp(1_700_000_001).expect("valid timestamp");
|
||||
let retention = OffsetDateTime::from_unix_timestamp(1_700_000_002).expect("valid timestamp");
|
||||
let legalhold = OffsetDateTime::from_unix_timestamp(1_700_000_003).expect("valid timestamp");
|
||||
opts.internal.tagging_timestamp = tagging;
|
||||
opts.internal.retention_timestamp = retention;
|
||||
opts.internal.legalhold_timestamp = legalhold;
|
||||
|
||||
let header = opts.header();
|
||||
for (suffix, expected) in [
|
||||
("source-replication-tagging-timestamp", tagging),
|
||||
("source-replication-retention-timestamp", retention),
|
||||
("source-replication-legalhold-timestamp", legalhold),
|
||||
] {
|
||||
assert_eq!(
|
||||
rustfs_utils::http::get_header(&header, suffix).as_deref(),
|
||||
Some(expected.format(&Rfc3339).expect("RFC3339 timestamp").as_str()),
|
||||
"replication put requests must carry the {suffix} header"
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn put_object_headers_omit_unset_replication_timestamps() {
|
||||
// UNIX_EPOCH means "never modified on the source"; sending it would
|
||||
// make the receiver treat an unset category as a fresh modification.
|
||||
let mut opts = PutObjectOptions::default();
|
||||
opts.internal.replication_request = true;
|
||||
opts.internal.tagging_timestamp = OffsetDateTime::UNIX_EPOCH;
|
||||
opts.internal.retention_timestamp = OffsetDateTime::UNIX_EPOCH;
|
||||
opts.internal.legalhold_timestamp = OffsetDateTime::UNIX_EPOCH;
|
||||
|
||||
let header = opts.header();
|
||||
for suffix in [
|
||||
"source-replication-tagging-timestamp",
|
||||
"source-replication-retention-timestamp",
|
||||
"source-replication-legalhold-timestamp",
|
||||
] {
|
||||
assert!(
|
||||
rustfs_utils::http::get_header(&header, suffix).is_none(),
|
||||
"unset {suffix} must not be sent to replication targets"
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn get_remote_target_client_internal_rejects_loopback_endpoint() {
|
||||
let sys = BucketTargetSys::default();
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -86,21 +86,6 @@ where
|
||||
com::save_config_with_opts(api, file, data, opts).await
|
||||
}
|
||||
|
||||
pub(crate) async fn save_config_with_opts_quiet<S>(api: Arc<S>, file: &str, data: Vec<u8>, opts: &ObjectOptions) -> Result<()>
|
||||
where
|
||||
S: ObjectIO<
|
||||
Error = Error,
|
||||
RangeSpec = HTTPRangeSpec,
|
||||
HeaderMap = HeaderMap,
|
||||
ObjectOptions = ObjectOptions,
|
||||
ObjectInfo = ObjectInfo,
|
||||
GetObjectReader = GetObjectReader,
|
||||
PutObjectReader = PutObjReader,
|
||||
>,
|
||||
{
|
||||
com::save_config_with_opts_quiet(api, file, data, opts).await
|
||||
}
|
||||
|
||||
pub(crate) async fn delete_config<S>(api: Arc<S>, file: &str) -> Result<()>
|
||||
where
|
||||
S: ObjectOperations<
|
||||
|
||||
@@ -45,104 +45,6 @@ const MANUAL_TRANSITION_JOB_LEASE_SECONDS: i128 = 60;
|
||||
const MANUAL_TRANSITION_LEGACY_SCOPE_SCAN_LIMIT: i32 = 1000;
|
||||
const MANUAL_TRANSITION_TASK_SCAN_LIMIT: i32 = 1000;
|
||||
const MANUAL_TRANSITION_WORKER_RESULT_SCAN_LIMIT: i32 = 1000;
|
||||
const MANUAL_TRANSITION_JOB_CAS_RETRIES: usize = 4;
|
||||
|
||||
#[cfg(test)]
|
||||
struct ManualTransitionJobCasBarrierState {
|
||||
job_id: Uuid,
|
||||
paused: std::sync::atomic::AtomicBool,
|
||||
arrived: tokio::sync::Notify,
|
||||
release: tokio::sync::Semaphore,
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
pub(crate) struct ManualTransitionJobCasBarrier {
|
||||
state: Arc<ManualTransitionJobCasBarrierState>,
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
static MANUAL_TRANSITION_JOB_CAS_BARRIER: std::sync::OnceLock<std::sync::Mutex<Option<Arc<ManualTransitionJobCasBarrierState>>>> =
|
||||
std::sync::OnceLock::new();
|
||||
|
||||
#[cfg(test)]
|
||||
impl ManualTransitionJobCasBarrier {
|
||||
pub(crate) fn install(job_id: Uuid) -> Self {
|
||||
let state = Arc::new(ManualTransitionJobCasBarrierState {
|
||||
job_id,
|
||||
paused: std::sync::atomic::AtomicBool::new(false),
|
||||
arrived: tokio::sync::Notify::new(),
|
||||
release: tokio::sync::Semaphore::new(0),
|
||||
});
|
||||
let mut slot = MANUAL_TRANSITION_JOB_CAS_BARRIER
|
||||
.get_or_init(|| std::sync::Mutex::new(None))
|
||||
.lock()
|
||||
.expect("manual transition progress CAS barrier mutex should not poison");
|
||||
assert!(
|
||||
slot.is_none(),
|
||||
"manual transition job CAS barrier must be installed by one test at a time"
|
||||
);
|
||||
*slot = Some(Arc::clone(&state));
|
||||
drop(slot);
|
||||
Self { state }
|
||||
}
|
||||
|
||||
pub(crate) async fn wait_until_paused(&self) {
|
||||
tokio::time::timeout(std::time::Duration::from_secs(30), async {
|
||||
loop {
|
||||
let arrived = self.state.arrived.notified();
|
||||
if self.state.paused.load(std::sync::atomic::Ordering::Acquire) {
|
||||
return;
|
||||
}
|
||||
arrived.await;
|
||||
}
|
||||
})
|
||||
.await
|
||||
.expect("manual transition job update should reach the deterministic CAS barrier");
|
||||
}
|
||||
|
||||
pub(crate) fn release(&self) {
|
||||
self.state.release.add_permits(1);
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
impl Drop for ManualTransitionJobCasBarrier {
|
||||
fn drop(&mut self) {
|
||||
self.release();
|
||||
let mut slot = MANUAL_TRANSITION_JOB_CAS_BARRIER
|
||||
.get_or_init(|| std::sync::Mutex::new(None))
|
||||
.lock()
|
||||
.expect("manual transition progress CAS barrier mutex should not poison");
|
||||
if slot.as_ref().is_some_and(|state| Arc::ptr_eq(state, &self.state)) {
|
||||
*slot = None;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
async fn pause_manual_transition_job_before_first_cas(job_id: Uuid) {
|
||||
let barrier = MANUAL_TRANSITION_JOB_CAS_BARRIER
|
||||
.get_or_init(|| std::sync::Mutex::new(None))
|
||||
.lock()
|
||||
.expect("manual transition progress CAS barrier mutex should not poison")
|
||||
.as_ref()
|
||||
.filter(|barrier| barrier.job_id == job_id)
|
||||
.cloned();
|
||||
if let Some(barrier) = barrier
|
||||
&& barrier
|
||||
.paused
|
||||
.compare_exchange(false, true, std::sync::atomic::Ordering::AcqRel, std::sync::atomic::Ordering::Acquire)
|
||||
.is_ok()
|
||||
{
|
||||
barrier.arrived.notify_one();
|
||||
barrier
|
||||
.release
|
||||
.acquire()
|
||||
.await
|
||||
.expect("manual transition job CAS barrier should remain open")
|
||||
.forget();
|
||||
}
|
||||
}
|
||||
|
||||
fn is_false(value: &bool) -> bool {
|
||||
!*value
|
||||
@@ -246,6 +148,7 @@ impl ManualTransitionJobRecord {
|
||||
|
||||
pub fn fail(&mut self, error: impl Into<String>) {
|
||||
self.state = ManualTransitionJobState::Failed;
|
||||
self.report.tier_failure = self.report.tier_failure.saturating_add(1);
|
||||
self.error = Some(error.into());
|
||||
self.mark_updated_terminal();
|
||||
}
|
||||
@@ -759,10 +662,6 @@ pub struct ManualTransitionWorkerResultRecord {
|
||||
}
|
||||
|
||||
impl ManualTransitionWorkerResultRecord {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub fn new(job_id: Uuid, task_key: impl Into<String>, result: ManualTransitionWorkerResult) -> Self {
|
||||
Self::new_with_reason(job_id, task_key, result, None)
|
||||
}
|
||||
@@ -1141,7 +1040,7 @@ pub async fn save_manual_transition_job_record_if_current(
|
||||
}
|
||||
let object = manual_transition_job_record_object_name(job.job_id).map_err(manual_transition_job_store_error)?;
|
||||
let data = job.encode().map_err(manual_transition_job_store_error)?;
|
||||
config_boundary::save_config_with_opts_quiet(
|
||||
config_boundary::save_config_with_opts(
|
||||
api,
|
||||
&object,
|
||||
data,
|
||||
@@ -1157,54 +1056,6 @@ pub async fn save_manual_transition_job_record_if_current(
|
||||
.await
|
||||
}
|
||||
|
||||
/// Applies a job-record mutation with optimistic concurrency control.
|
||||
///
|
||||
/// The mutation returns whether the record needs to be persisted. When a lease
|
||||
/// is supplied, ownership is checked again after every conflicting write.
|
||||
pub async fn update_manual_transition_job_record<F>(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Option<Uuid>,
|
||||
update: F,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord>
|
||||
where
|
||||
F: FnMut(&mut ManualTransitionJobRecord) -> bool,
|
||||
{
|
||||
update_manual_transition_job_record_from(api, job_id, expected_lease_id, None, update).await
|
||||
}
|
||||
|
||||
async fn update_manual_transition_job_record_from<F>(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Option<Uuid>,
|
||||
mut current: Option<(ManualTransitionJobRecord, String)>,
|
||||
mut update: F,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord>
|
||||
where
|
||||
F: FnMut(&mut ManualTransitionJobRecord) -> bool,
|
||||
{
|
||||
for _ in 0..MANUAL_TRANSITION_JOB_CAS_RETRIES {
|
||||
let (mut record, etag) = match current.take() {
|
||||
Some(current) => current,
|
||||
None => load_manual_transition_job_record_with_etag(api.clone(), job_id).await?,
|
||||
};
|
||||
if expected_lease_id.is_some_and(|lease_id| record.lease_id != lease_id) {
|
||||
return Err(Error::PreconditionFailed);
|
||||
}
|
||||
if !update(&mut record) {
|
||||
return Ok(record);
|
||||
}
|
||||
#[cfg(test)]
|
||||
pause_manual_transition_job_before_first_cas(job_id).await;
|
||||
match save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await {
|
||||
Ok(()) => return Ok(record),
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
Err(Error::PreconditionFailed)
|
||||
}
|
||||
|
||||
pub(crate) async fn save_manual_transition_worker_result_if_absent(
|
||||
api: Arc<ECStore>,
|
||||
record: &ManualTransitionWorkerResultRecord,
|
||||
@@ -1261,10 +1112,6 @@ pub(crate) async fn save_manual_transition_task_if_absent(
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn load_manual_transition_task_record(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
@@ -1328,10 +1175,6 @@ async fn scan_manual_transition_task_journal(api: Arc<ECStore>, job_id: Uuid) ->
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn load_manual_transition_worker_result_stats(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
@@ -1467,121 +1310,103 @@ async fn scan_manual_transition_worker_result_journal(
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn reconcile_manual_transition_worker_results(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
reconcile_manual_transition_worker_results_inner(api, job_id, None, queue_snapshot, false).await
|
||||
}
|
||||
|
||||
pub(crate) async fn reconcile_manual_transition_worker_results_if_owned(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Uuid,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
reconcile_manual_transition_worker_results_inner(api, job_id, Some(expected_lease_id), queue_snapshot, false).await
|
||||
}
|
||||
|
||||
async fn reconcile_manual_transition_worker_results_inner(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Option<Uuid>,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
mark_missing_results_unknown: bool,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let task_stats = match scan_manual_transition_task_journal(api.clone(), job_id).await? {
|
||||
ManualTransitionTaskJournal::Stats(stats) => stats,
|
||||
ManualTransitionTaskJournal::Corrupt(error) => {
|
||||
return mark_manual_transition_job_unknown_for_task_journal_error(
|
||||
api,
|
||||
job_id,
|
||||
expected_lease_id,
|
||||
error,
|
||||
queue_snapshot,
|
||||
)
|
||||
.await;
|
||||
return mark_manual_transition_job_unknown_for_task_journal_error(api, job_id, error, queue_snapshot).await;
|
||||
}
|
||||
};
|
||||
let stats = match scan_manual_transition_worker_result_journal(api.clone(), job_id).await? {
|
||||
ManualTransitionWorkerResultJournal::Stats(stats) => stats,
|
||||
ManualTransitionWorkerResultJournal::Corrupt(error) => {
|
||||
return mark_manual_transition_job_unknown_for_worker_result_journal_error(
|
||||
api,
|
||||
job_id,
|
||||
expected_lease_id,
|
||||
error,
|
||||
queue_snapshot,
|
||||
)
|
||||
.await;
|
||||
return mark_manual_transition_job_unknown_for_worker_result_journal_error(api, job_id, error, queue_snapshot).await;
|
||||
}
|
||||
};
|
||||
let mut changed = false;
|
||||
let record = update_manual_transition_job_record(api.clone(), job_id, expected_lease_id, |record| {
|
||||
let counts_changed = record.apply_worker_result_counts(
|
||||
for _ in 0..4 {
|
||||
let (mut record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
let changed = record.apply_worker_result_counts(
|
||||
stats.stats.completed,
|
||||
stats.stats.failed,
|
||||
&stats.stats.tier_failure_by_reason,
|
||||
task_stats.queued,
|
||||
queue_snapshot,
|
||||
);
|
||||
let became_unknown = mark_missing_results_unknown && record.mark_unknown_if_worker_results_lost(queue_snapshot);
|
||||
changed = counts_changed || became_unknown;
|
||||
changed
|
||||
})
|
||||
.await?;
|
||||
if !changed {
|
||||
return Ok(record);
|
||||
if !changed {
|
||||
return Ok(record);
|
||||
}
|
||||
match save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await {
|
||||
Ok(()) => {
|
||||
if record.is_terminal() {
|
||||
delete_manual_transition_scope_admission_if_current(
|
||||
api.clone(),
|
||||
&record.scope_key,
|
||||
record.job_id,
|
||||
record.lease_id,
|
||||
)
|
||||
.await?;
|
||||
} else {
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
}
|
||||
return Ok(record);
|
||||
}
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
if record.is_terminal() {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id).await?;
|
||||
} else {
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
}
|
||||
Ok(record)
|
||||
Err(Error::PreconditionFailed)
|
||||
}
|
||||
|
||||
async fn mark_manual_transition_job_unknown_for_task_journal_error(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Option<Uuid>,
|
||||
error: String,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let mut changed = false;
|
||||
let record = update_manual_transition_job_record(api.clone(), job_id, expected_lease_id, |record| {
|
||||
changed = record.mark_unknown_for_task_journal_error(error.clone(), queue_snapshot);
|
||||
changed
|
||||
})
|
||||
.await?;
|
||||
if changed && record.is_terminal() {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id).await?;
|
||||
for _ in 0..4 {
|
||||
let (mut record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
if !record.mark_unknown_for_task_journal_error(error.clone(), queue_snapshot) {
|
||||
return Ok(record);
|
||||
}
|
||||
match save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await {
|
||||
Ok(()) => {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id)
|
||||
.await?;
|
||||
return Ok(record);
|
||||
}
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
Ok(record)
|
||||
Err(Error::PreconditionFailed)
|
||||
}
|
||||
|
||||
async fn mark_manual_transition_job_unknown_for_worker_result_journal_error(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Option<Uuid>,
|
||||
error: String,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let mut changed = false;
|
||||
let record = update_manual_transition_job_record(api.clone(), job_id, expected_lease_id, |record| {
|
||||
changed = record.mark_unknown_for_worker_result_journal_error(error.clone(), queue_snapshot);
|
||||
changed
|
||||
})
|
||||
.await?;
|
||||
if changed && record.is_terminal() {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id).await?;
|
||||
for _ in 0..4 {
|
||||
let (mut record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
if !record.mark_unknown_for_worker_result_journal_error(error.clone(), queue_snapshot) {
|
||||
return Ok(record);
|
||||
}
|
||||
match save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await {
|
||||
Ok(()) => {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id)
|
||||
.await?;
|
||||
return Ok(record);
|
||||
}
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
Ok(record)
|
||||
Err(Error::PreconditionFailed)
|
||||
}
|
||||
|
||||
pub async fn save_manual_transition_scope_admission_if_absent(
|
||||
@@ -1778,14 +1603,19 @@ async fn find_active_legacy_manual_transition_scope_conflict(
|
||||
}
|
||||
|
||||
pub async fn request_manual_transition_job_cancel(api: Arc<ECStore>, job_id: Uuid) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
update_manual_transition_job_record(api, job_id, None, |record| {
|
||||
for _ in 0..4 {
|
||||
let (mut record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
if record.is_terminal() || record.cancel_requested {
|
||||
return false;
|
||||
return Ok(record);
|
||||
}
|
||||
record.mark_cancel_requested();
|
||||
true
|
||||
})
|
||||
.await
|
||||
match save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await {
|
||||
Ok(()) => return Ok(record),
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
Err(Error::PreconditionFailed)
|
||||
}
|
||||
|
||||
pub async fn persist_manual_transition_job_progress(
|
||||
@@ -1794,39 +1624,10 @@ pub async fn persist_manual_transition_job_progress(
|
||||
report: &ManualTransitionRunReport,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let current = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
persist_manual_transition_job_progress_inner(api, job_id, current.0.lease_id, Some(current), report, queue_snapshot).await
|
||||
}
|
||||
|
||||
pub async fn persist_manual_transition_job_progress_if_owned(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Uuid,
|
||||
report: &ManualTransitionRunReport,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
persist_manual_transition_job_progress_inner(api, job_id, expected_lease_id, None, report, queue_snapshot).await
|
||||
}
|
||||
|
||||
async fn persist_manual_transition_job_progress_inner(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Uuid,
|
||||
current: Option<(ManualTransitionJobRecord, String)>,
|
||||
report: &ManualTransitionRunReport,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let record = update_manual_transition_job_record_from(api.clone(), job_id, Some(expected_lease_id), current, |record| {
|
||||
if record.state != ManualTransitionJobState::Running {
|
||||
return false;
|
||||
}
|
||||
record.update_running_progress(report.clone(), queue_snapshot);
|
||||
true
|
||||
})
|
||||
.await?;
|
||||
if record.state == ManualTransitionJobState::Running {
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
}
|
||||
let (mut record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
record.update_running_progress(report.clone(), queue_snapshot);
|
||||
save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await?;
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
Ok(record)
|
||||
}
|
||||
|
||||
@@ -1860,58 +1661,25 @@ pub async fn renew_manual_transition_job_lease(
|
||||
job_id: Uuid,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let current = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
renew_manual_transition_job_lease_inner(api, job_id, current.0.lease_id, Some(current), queue_snapshot).await
|
||||
}
|
||||
|
||||
pub async fn renew_manual_transition_job_lease_if_owned(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Uuid,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
renew_manual_transition_job_lease_inner(api, job_id, expected_lease_id, None, queue_snapshot).await
|
||||
}
|
||||
|
||||
async fn renew_manual_transition_job_lease_inner(
|
||||
api: Arc<ECStore>,
|
||||
job_id: Uuid,
|
||||
expected_lease_id: Uuid,
|
||||
current: Option<(ManualTransitionJobRecord, String)>,
|
||||
queue_snapshot: ManualTransitionQueueSnapshot,
|
||||
) -> EcstoreResult<ManualTransitionJobRecord> {
|
||||
let (current, current_etag) = match current {
|
||||
Some(current) => current,
|
||||
None => load_manual_transition_job_record_with_etag(api.clone(), job_id).await?,
|
||||
};
|
||||
if current.lease_id != expected_lease_id {
|
||||
return Err(Error::PreconditionFailed);
|
||||
}
|
||||
if current.state != ManualTransitionJobState::Running {
|
||||
return Ok(current);
|
||||
}
|
||||
if current.scan_completed && queue_snapshot.queued == 0 && queue_snapshot.active == 0 {
|
||||
return reconcile_manual_transition_worker_results_inner(api, job_id, Some(expected_lease_id), queue_snapshot, true)
|
||||
.await;
|
||||
}
|
||||
let record = update_manual_transition_job_record_from(
|
||||
api.clone(),
|
||||
job_id,
|
||||
Some(expected_lease_id),
|
||||
Some((current, current_etag)),
|
||||
|record| {
|
||||
if record.state != ManualTransitionJobState::Running {
|
||||
return false;
|
||||
let (mut record, mut etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
if record.state == ManualTransitionJobState::Running {
|
||||
if record.scan_completed && queue_snapshot.queued == 0 && queue_snapshot.active == 0 {
|
||||
record = reconcile_manual_transition_worker_results(api.clone(), job_id, queue_snapshot).await?;
|
||||
if record.is_terminal() || !record.report.worker_transition_pending() {
|
||||
return Ok(record);
|
||||
}
|
||||
(record, etag) = load_manual_transition_job_record_with_etag(api.clone(), job_id).await?;
|
||||
}
|
||||
let became_terminal = record.mark_unknown_if_worker_results_lost(queue_snapshot);
|
||||
if !became_terminal {
|
||||
record.renew_lease(queue_snapshot);
|
||||
true
|
||||
},
|
||||
)
|
||||
.await?;
|
||||
if record.is_terminal() {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id).await?;
|
||||
} else if record.state == ManualTransitionJobState::Running {
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
}
|
||||
save_manual_transition_job_record_if_current(api.clone(), &record, &etag).await?;
|
||||
if became_terminal {
|
||||
delete_manual_transition_scope_admission_if_current(api, &record.scope_key, record.job_id, record.lease_id).await?;
|
||||
} else {
|
||||
renew_manual_transition_scope_admission_from_job(api, &record).await?;
|
||||
}
|
||||
}
|
||||
Ok(record)
|
||||
}
|
||||
@@ -1920,31 +1688,15 @@ async fn renew_manual_transition_scope_admission_from_job(
|
||||
api: Arc<ECStore>,
|
||||
record: &ManualTransitionJobRecord,
|
||||
) -> EcstoreResult<()> {
|
||||
for _ in 0..MANUAL_TRANSITION_JOB_CAS_RETRIES {
|
||||
let (admission, admission_etag) =
|
||||
match load_manual_transition_scope_admission_with_etag(api.clone(), &record.scope_key).await {
|
||||
Ok(admission) => admission,
|
||||
Err(Error::ConfigNotFound) => return Ok(()),
|
||||
Err(err) => return Err(err),
|
||||
};
|
||||
if admission.job_id != record.job_id || admission.lease_id != record.lease_id {
|
||||
return Err(Error::PreconditionFailed);
|
||||
}
|
||||
let mut renewed_admission = ManualTransitionScopeAdmission::from_job(record);
|
||||
renewed_admission.lease_expires_at_unix_nanos = renewed_admission
|
||||
.lease_expires_at_unix_nanos
|
||||
.max(admission.lease_expires_at_unix_nanos);
|
||||
renewed_admission.updated_at_unix_nanos = renewed_admission.updated_at_unix_nanos.max(admission.updated_at_unix_nanos);
|
||||
if renewed_admission == admission {
|
||||
return Ok(());
|
||||
}
|
||||
match save_manual_transition_scope_admission_if_current(api.clone(), &renewed_admission, &admission_etag).await {
|
||||
Ok(()) => return Ok(()),
|
||||
Err(Error::PreconditionFailed) => continue,
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
if let Ok((admission, admission_etag)) =
|
||||
load_manual_transition_scope_admission_with_etag(api.clone(), &record.scope_key).await
|
||||
&& admission.job_id == record.job_id
|
||||
&& admission.lease_id == record.lease_id
|
||||
{
|
||||
let renewed_admission = ManualTransitionScopeAdmission::from_job(record);
|
||||
save_manual_transition_scope_admission_if_current(api, &renewed_admission, &admission_etag).await?;
|
||||
}
|
||||
Err(Error::PreconditionFailed)
|
||||
Ok(())
|
||||
}
|
||||
|
||||
pub async fn delete_manual_transition_scope_admission_if_current(
|
||||
@@ -2634,14 +2386,14 @@ mod tests {
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn manual_transition_job_record_control_plane_failure_does_not_count_tier_failure() {
|
||||
fn manual_transition_job_record_failure_counts_tier_failure() {
|
||||
let options = ManualTransitionRunOptions::default();
|
||||
let mut record = ManualTransitionJobRecord::new(Uuid::new_v4(), "bucket", &options, TEST_OWNER);
|
||||
|
||||
record.fail("missing tier");
|
||||
|
||||
assert_eq!(record.state, ManualTransitionJobState::Failed);
|
||||
assert_eq!(record.report.tier_failure, 0);
|
||||
assert_eq!(record.report.tier_failure, 1);
|
||||
assert_eq!(record.error.as_deref(), Some("missing tier"));
|
||||
}
|
||||
|
||||
|
||||
@@ -15,35 +15,25 @@
|
||||
use rustfs_common::metrics::IlmAction;
|
||||
|
||||
use crate::bucket::lifecycle::lifecycle::ObjectOpts;
|
||||
use crate::bucket::replication::ReplicationLifecycleBridge;
|
||||
pub(crate) use crate::bucket::replication::ReplicationStatusType;
|
||||
#[cfg(test)]
|
||||
pub(crate) use crate::bucket::replication::VersionPurgeStatusType;
|
||||
pub(crate) use crate::bucket::replication::{
|
||||
DeleteReplicationConfigSnapshot, ReplicationObjectBridge, replication_state_to_filemeta,
|
||||
};
|
||||
use crate::bucket::replication::{ReplicationLifecycleBridge, ReplicationLifecycleConfig};
|
||||
use crate::storage_api_contracts::object::DeletedObject;
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) type LifecycleReplicationConfig = ReplicationLifecycleConfig;
|
||||
|
||||
pub(crate) fn has_pending_version_purge(obj: &ObjectOpts) -> bool {
|
||||
obj.version_purge_status.is_pending()
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn has_pending_object_replication(obj: &ObjectOpts) -> bool {
|
||||
replication_status_blocks_lifecycle(&obj.replication_status)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn has_pending_lifecycle_replication(obj: &ObjectOpts) -> bool {
|
||||
has_pending_object_replication(obj) || has_pending_version_purge(obj)
|
||||
}
|
||||
|
||||
@@ -14,10 +14,6 @@
|
||||
|
||||
use std::collections::HashMap;
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn decode_tags_to_map(tags: &str) -> HashMap<String, String> {
|
||||
crate::bucket::tagging::decode_tags_to_map(tags)
|
||||
}
|
||||
|
||||
@@ -331,10 +331,6 @@ where
|
||||
persist_tier_delete_journal_entry(api, &committed).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn abort_tier_delete_journal_entry<S>(api: Arc<S>, je: &Jentry) -> std::io::Result<()>
|
||||
where
|
||||
S: ObjectOperations<
|
||||
|
||||
@@ -148,10 +148,6 @@ struct RecoveryCursor {
|
||||
object: String,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn recover_tier_free_versions(
|
||||
api: Arc<ECStore>,
|
||||
limit: usize,
|
||||
|
||||
@@ -385,10 +385,6 @@ impl ExpiryOp for Jentry {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn delete_object_from_remote_tier(obj_name: &str, rv_id: &str, tier_name: &str) -> Result<(), std::io::Error> {
|
||||
let result = delete_object_from_remote_tier_raw(obj_name, rv_id, tier_name).await;
|
||||
if let Err(err) = &result
|
||||
@@ -399,10 +395,6 @@ pub async fn delete_object_from_remote_tier(obj_name: &str, rv_id: &str, tier_na
|
||||
result
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
async fn delete_object_from_remote_tier_raw(obj_name: &str, rv_id: &str, tier_name: &str) -> Result<(), std::io::Error> {
|
||||
#[cfg(test)]
|
||||
if let Some(result) = run_remote_tier_delete_test_hook(obj_name, rv_id, tier_name) {
|
||||
@@ -413,10 +405,6 @@ async fn delete_object_from_remote_tier_raw(obj_name: &str, rv_id: &str, tier_na
|
||||
delete_object_from_remote_tier_raw_with_manager(obj_name, rv_id, tier_name, &tier_config_mgr).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
async fn delete_object_from_remote_tier_raw_with_manager(
|
||||
obj_name: &str,
|
||||
rv_id: &str,
|
||||
@@ -497,10 +485,6 @@ pub enum RemoteTierDeleteOutcome {
|
||||
AlreadyRemoved,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
pub async fn delete_object_from_remote_tier_idempotent(
|
||||
obj_name: &str,
|
||||
rv_id: &str,
|
||||
|
||||
@@ -50,16 +50,8 @@ pub type Result<T> = std::result::Result<T, TransitionTransactionError>;
|
||||
#[derive(Debug, thiserror::Error)]
|
||||
pub enum TransitionTransactionError {
|
||||
#[error("transition transaction already exists")]
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
AlreadyExists,
|
||||
#[error("transition transaction is not found")]
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity tier/lifecycle entry point that this port never wired (backlog#1823)"
|
||||
)]
|
||||
NotFound,
|
||||
#[error("transition transaction is corrupt: {0}")]
|
||||
Corrupt(&'static str),
|
||||
|
||||
@@ -50,75 +50,6 @@ use uuid::Uuid;
|
||||
|
||||
const BUCKET_METADATA_REFRESH_INTERVAL: Duration = Duration::from_secs(15 * 60);
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
struct ConfigWriteLockProbeState {
|
||||
bucket: String,
|
||||
arrived: tokio::sync::Notify,
|
||||
}
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
static CONFIG_WRITE_LOCK_PROBES: std::sync::OnceLock<StdMutex<Vec<Arc<ConfigWriteLockProbeState>>>> = std::sync::OnceLock::new();
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
#[allow(dead_code, reason = "installed by tests behind `--features test-util` (backlog#1823)")]
|
||||
pub struct ConfigWriteLockProbe {
|
||||
state: Arc<ConfigWriteLockProbeState>,
|
||||
}
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
impl ConfigWriteLockProbe {
|
||||
#[allow(dead_code, reason = "installed by tests behind `--features test-util` (backlog#1823)")]
|
||||
pub fn install(bucket: &str) -> Self {
|
||||
let state = Arc::new(ConfigWriteLockProbeState {
|
||||
bucket: bucket.to_string(),
|
||||
arrived: tokio::sync::Notify::new(),
|
||||
});
|
||||
let mut probes = CONFIG_WRITE_LOCK_PROBES
|
||||
.get_or_init(|| StdMutex::new(Vec::new()))
|
||||
.lock()
|
||||
.expect("config write lock probe mutex should not poison");
|
||||
assert!(
|
||||
!probes.iter().any(|current| current.bucket == state.bucket),
|
||||
"config write lock probe must be unique for a bucket"
|
||||
);
|
||||
probes.push(Arc::clone(&state));
|
||||
drop(probes);
|
||||
Self { state }
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "installed by tests behind `--features test-util` (backlog#1823)")]
|
||||
pub async fn wait_until_attempted(&self) {
|
||||
tokio::time::timeout(Duration::from_secs(30), self.state.arrived.notified())
|
||||
.await
|
||||
.expect("bucket config update should attempt the transaction lock");
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
impl Drop for ConfigWriteLockProbe {
|
||||
fn drop(&mut self) {
|
||||
let mut probes = CONFIG_WRITE_LOCK_PROBES
|
||||
.get_or_init(|| StdMutex::new(Vec::new()))
|
||||
.lock()
|
||||
.expect("config write lock probe mutex should not poison");
|
||||
probes.retain(|state| !Arc::ptr_eq(state, &self.state));
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
fn notify_config_write_lock_attempt(bucket: &str) {
|
||||
let probe = CONFIG_WRITE_LOCK_PROBES
|
||||
.get_or_init(|| StdMutex::new(Vec::new()))
|
||||
.lock()
|
||||
.expect("config write lock probe mutex should not poison")
|
||||
.iter()
|
||||
.find(|probe| probe.bucket == bucket)
|
||||
.cloned();
|
||||
if let Some(probe) = probe {
|
||||
probe.arrived.notify_one();
|
||||
}
|
||||
}
|
||||
|
||||
#[derive(Clone, Copy)]
|
||||
enum MetadataLoadMode {
|
||||
Initial,
|
||||
@@ -659,41 +590,6 @@ pub async fn update_under_transaction_lock(
|
||||
update_under_config_write_guard(get_bucket_metadata_sys()?, guard, config_file, data).await
|
||||
}
|
||||
|
||||
/// Clear one config file while the caller holds this bucket's transaction lock.
|
||||
pub async fn delete_under_transaction_lock(
|
||||
guard: &BucketMetadataMutationGuard,
|
||||
bucket: &str,
|
||||
config_file: &str,
|
||||
) -> Result<OffsetDateTime> {
|
||||
guard.ensure_valid(bucket)?;
|
||||
delete_under_config_write_guard(get_bucket_metadata_sys()?, guard, config_file).await
|
||||
}
|
||||
|
||||
pub async fn update_quota_if_incarnation(
|
||||
bucket: &str,
|
||||
data: Vec<u8>,
|
||||
expected_incarnation_id: Uuid,
|
||||
proof: &crate::services::notification_sys::CrossPoolFenceFleetProofToken,
|
||||
) -> Result<OffsetDateTime> {
|
||||
let sys = get_bucket_metadata_sys()?;
|
||||
let guard = Box::pin(acquire_config_write_guard_for_incarnation(
|
||||
sys.clone(),
|
||||
bucket,
|
||||
Some(expected_incarnation_id),
|
||||
))
|
||||
.await?;
|
||||
if !crate::services::notification_sys::cross_pool_fence_fleet_proof_matches(proof) {
|
||||
return Err(Error::NamespaceLockQuorumUnavailable {
|
||||
mode: "quota_capability",
|
||||
bucket: bucket.to_string(),
|
||||
object: rustfs_config::QUOTA_CONFIG_FILE.to_string(),
|
||||
required: 1,
|
||||
achieved: 0,
|
||||
});
|
||||
}
|
||||
update_under_config_write_guard(sys, &guard, rustfs_config::QUOTA_CONFIG_FILE, data).await
|
||||
}
|
||||
|
||||
pub async fn update_bucket_targets_under_transaction_lock(
|
||||
guard: &BucketMetadataMutationGuard,
|
||||
bucket: &str,
|
||||
@@ -808,14 +704,6 @@ pub async fn acquire_bucket_metadata_transaction_lock(bucket: &str) -> Result<Bu
|
||||
acquire_config_write_guard(get_bucket_metadata_sys()?, bucket).await
|
||||
}
|
||||
|
||||
/// Acquire the bucket transaction lock only if its incarnation still matches.
|
||||
pub async fn acquire_bucket_metadata_transaction_lock_for_incarnation(
|
||||
bucket: &str,
|
||||
expected_incarnation_id: Uuid,
|
||||
) -> Result<BucketMetadataMutationGuard> {
|
||||
acquire_config_write_guard_for_incarnation(get_bucket_metadata_sys()?, bucket, Some(expected_incarnation_id)).await
|
||||
}
|
||||
|
||||
pub(crate) async fn acquire_bucket_metadata_transaction_lock_in(
|
||||
ctx: &crate::runtime::instance::InstanceContext,
|
||||
bucket: &str,
|
||||
@@ -846,26 +734,7 @@ async fn acquire_transaction_lock_with_sys(
|
||||
let lock = api
|
||||
.new_ns_lock(RUSTFS_META_BUCKET, &bucket_metadata_transaction_lock_key(bucket))
|
||||
.await?;
|
||||
let acquire = lock.get_write_lock(crate::set_disk::get_lock_acquire_timeout());
|
||||
#[cfg(any(test, feature = "test-util"))]
|
||||
{
|
||||
tokio::pin!(acquire);
|
||||
let mut notified = false;
|
||||
let guard = futures::future::poll_fn(|cx| match std::future::Future::poll(acquire.as_mut(), cx) {
|
||||
std::task::Poll::Pending => {
|
||||
if !notified {
|
||||
notify_config_write_lock_attempt(bucket);
|
||||
notified = true;
|
||||
}
|
||||
std::task::Poll::Pending
|
||||
}
|
||||
std::task::Poll::Ready(result) => std::task::Poll::Ready(result),
|
||||
})
|
||||
.await?;
|
||||
Ok(guard)
|
||||
}
|
||||
#[cfg(not(any(test, feature = "test-util")))]
|
||||
Ok(acquire.await?)
|
||||
Ok(lock.get_write_lock(crate::set_disk::get_lock_acquire_timeout()).await?)
|
||||
}
|
||||
|
||||
/// The lock resource name is deliberately still the `bucket-targets` one it
|
||||
@@ -893,10 +762,6 @@ pub async fn get_bucket_policy_raw(bucket: &str) -> Result<(String, OffsetDateTi
|
||||
bucket_meta_sys.get_bucket_policy_raw(bucket).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "free-function facade over the live BucketMetadataSys::get_bucket_acl_config; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub async fn get_bucket_acl_config(bucket: &str) -> Result<(String, OffsetDateTime)> {
|
||||
let bucket_meta_sys_lock = get_bucket_metadata_sys()?;
|
||||
let bucket_meta_sys = bucket_meta_sys_lock.read().await;
|
||||
@@ -1024,37 +889,6 @@ pub(crate) async fn get_object_lock_config_and_incarnation_from_disk_in(
|
||||
}
|
||||
}
|
||||
|
||||
/// Re-read the quota configuration and bucket incarnation from the same
|
||||
/// authoritative metadata blob while the caller holds the bucket metadata
|
||||
/// transaction read lock.
|
||||
pub(crate) async fn get_quota_config_and_incarnation_from_disk_in(
|
||||
ctx: &crate::runtime::instance::InstanceContext,
|
||||
bucket: &str,
|
||||
) -> Result<(Option<BucketQuota>, Uuid, OffsetDateTime)> {
|
||||
let bucket_meta_sys_lock = bucket_metadata_sys_of(ctx)?;
|
||||
let bucket_meta_sys = bucket_meta_sys_lock.read().await.clone();
|
||||
|
||||
match bucket_meta_sys
|
||||
.read_authoritative_metadata_from_disk_under_transaction_lock(bucket)
|
||||
.await?
|
||||
{
|
||||
BucketMetadataAuthority::Authoritative(metadata)
|
||||
if metadata.bucket_incarnation_sidecar && !metadata.bucket_incarnation_id.is_nil() =>
|
||||
{
|
||||
Ok((
|
||||
metadata.quota_config.clone(),
|
||||
metadata.bucket_incarnation_id,
|
||||
metadata.quota_config_updated_at,
|
||||
))
|
||||
}
|
||||
BucketMetadataAuthority::Authoritative(_) => {
|
||||
Err(Error::other(format!("bucket incarnation metadata is not authoritative: {bucket}")))
|
||||
}
|
||||
BucketMetadataAuthority::MissingBucket => Err(Error::BucketNotFound(bucket.to_string())),
|
||||
BucketMetadataAuthority::Fabricated => Err(Error::other(format!("bucket quota metadata is not authoritative: {bucket}"))),
|
||||
}
|
||||
}
|
||||
|
||||
pub async fn get_replication_config(bucket: &str) -> Result<(ReplicationConfiguration, OffsetDateTime)> {
|
||||
let bucket_meta_sys_lock = get_bucket_metadata_sys()?;
|
||||
let bucket_meta_sys = bucket_meta_sys_lock.read().await;
|
||||
@@ -1111,10 +945,6 @@ pub async fn get_config_from_disk(bucket: &str) -> Result<BucketMetadata> {
|
||||
bucket_meta_sys.get_config_from_disk(bucket).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "ambient-facade variant of the live created_at_in; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub async fn created_at(bucket: &str) -> Result<OffsetDateTime> {
|
||||
let bucket_meta_sys_lock = get_bucket_metadata_sys()?;
|
||||
let bucket_meta_sys = bucket_meta_sys_lock.read().await;
|
||||
@@ -1628,7 +1458,6 @@ impl BucketMetadataSys {
|
||||
/// [`Self::update`], with the payload computed from the loaded metadata
|
||||
/// instead of supplied up front. Loads through this system's own store so
|
||||
/// the read and the persisted write target the same instance.
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
async fn update_config_with<F>(&self, bucket: &str, config_file: &str, mutate: F) -> Result<OffsetDateTime>
|
||||
where
|
||||
F: FnOnce(&BucketMetadata) -> Result<Vec<u8>> + Send,
|
||||
@@ -1733,7 +1562,6 @@ impl BucketMetadataSys {
|
||||
/// A miss is never published as an authoritative default, and a snapshot
|
||||
/// read before delete plus same-name recreation cannot replace the new
|
||||
/// generation.
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub(crate) async fn reload_from_store(&self, bucket: &str) -> Result<()> {
|
||||
if is_meta_bucketname(bucket) {
|
||||
return Err(Error::other("errInvalidArgument"));
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: bucket subsystems still contain staged ECStore migration code.
|
||||
#![allow(dead_code)]
|
||||
|
||||
pub mod bandwidth;
|
||||
pub mod bucket_target_sys;
|
||||
|
||||
@@ -136,7 +136,6 @@ pub fn add_years(dt: OffsetDateTime, years: i32) -> OffsetDateTime {
|
||||
|
||||
/// Check if an object has legal hold enabled.
|
||||
/// Returns true if legal hold is ON.
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
fn has_legal_hold(user_defined: &std::collections::HashMap<String, String>) -> bool {
|
||||
let lhold = objectlock::get_object_legalhold_meta(user_defined);
|
||||
matches!(lhold.status, Some(ref st) if st.as_str() == ObjectLockLegalHoldStatus::ON)
|
||||
@@ -152,7 +151,6 @@ fn has_legal_hold(user_defined: &std::collections::HashMap<String, String>) -> b
|
||||
/// # Returns
|
||||
/// * `true` if the object is locked (cannot be deleted/modified)
|
||||
/// * `false` if the object is not locked
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn is_object_locked_by_metadata(user_defined: &std::collections::HashMap<String, String>, is_delete_marker: bool) -> bool {
|
||||
// Delete markers are never locked
|
||||
if is_delete_marker {
|
||||
|
||||
@@ -52,7 +52,6 @@ impl QuotaChecker {
|
||||
) -> Result<QuotaCheckResult, QuotaError> {
|
||||
let start_time = Instant::now();
|
||||
let quota_config = self.get_quota_config(bucket).await?;
|
||||
let uses_durable_reservations = quota_config.uses_durable_reservations();
|
||||
|
||||
// If no quota limit is set, allow operation
|
||||
let quota_limit = match quota_config.quota {
|
||||
@@ -68,7 +67,6 @@ impl QuotaChecker {
|
||||
quota_limit: None,
|
||||
operation_size,
|
||||
remaining: None,
|
||||
uses_durable_reservations,
|
||||
});
|
||||
}
|
||||
Some(q) => q,
|
||||
@@ -76,17 +74,14 @@ impl QuotaChecker {
|
||||
|
||||
let current_usage = self.get_real_time_usage(bucket).await?;
|
||||
|
||||
let admission_size = if uses_durable_reservations { 0 } else { operation_size };
|
||||
let expected_usage = match operation {
|
||||
QuotaOperation::PutObject | QuotaOperation::PostObject | QuotaOperation::CopyObject => {
|
||||
current_usage.saturating_add(admission_size)
|
||||
}
|
||||
QuotaOperation::PutObject | QuotaOperation::PostObject | QuotaOperation::CopyObject => current_usage + operation_size,
|
||||
QuotaOperation::DeleteObject => current_usage.saturating_sub(operation_size),
|
||||
};
|
||||
|
||||
let allowed = match operation {
|
||||
QuotaOperation::PutObject | QuotaOperation::PostObject | QuotaOperation::CopyObject => {
|
||||
quota_config.check_operation_allowed(current_usage, admission_size)
|
||||
quota_config.check_operation_allowed(current_usage, operation_size)
|
||||
}
|
||||
QuotaOperation::DeleteObject => true,
|
||||
};
|
||||
@@ -110,7 +105,6 @@ impl QuotaChecker {
|
||||
quota_limit: Some(quota_limit),
|
||||
operation_size,
|
||||
remaining,
|
||||
uses_durable_reservations,
|
||||
};
|
||||
|
||||
let duration = start_time.elapsed();
|
||||
@@ -164,26 +158,6 @@ impl QuotaChecker {
|
||||
.await
|
||||
}
|
||||
|
||||
pub async fn set_durable_quota_config_if_incarnation(
|
||||
&mut self,
|
||||
bucket: &str,
|
||||
quota: BucketQuota,
|
||||
expected_incarnation_id: uuid::Uuid,
|
||||
proof: &crate::services::notification_sys::CrossPoolFenceFleetProofToken,
|
||||
) -> Result<OffsetDateTime, QuotaError> {
|
||||
let json_data = serde_json::to_vec("a).map_err(|e| QuotaError::InvalidConfig {
|
||||
reason: format!("Failed to serialize quota config: {}", e),
|
||||
})?;
|
||||
let start_time = Instant::now();
|
||||
let updated_at =
|
||||
crate::bucket::metadata_sys::update_quota_if_incarnation(bucket, json_data, expected_incarnation_id, proof)
|
||||
.await
|
||||
.map_err(QuotaError::StorageError)?;
|
||||
|
||||
rustfs_common::metrics::Metrics::inc_time(Metric::QuotaSync, start_time.elapsed());
|
||||
Ok(updated_at)
|
||||
}
|
||||
|
||||
async fn set_quota_config_for_incarnation(
|
||||
&mut self,
|
||||
bucket: &str,
|
||||
@@ -381,7 +355,6 @@ mod tests {
|
||||
quota_limit: None,
|
||||
operation_size: 1024,
|
||||
remaining: None,
|
||||
uses_durable_reservations: false,
|
||||
};
|
||||
|
||||
assert!(result.allowed);
|
||||
@@ -405,13 +378,4 @@ mod tests {
|
||||
let allowed = quota.check_operation_allowed(512, 1024);
|
||||
assert!(!allowed);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn legacy_quota_rejects_full_operation_while_v1_defers_net_growth() {
|
||||
let legacy: BucketQuota = serde_json::from_str(r#"{"quota":5}"#).expect("legacy quota should parse");
|
||||
let durable = BucketQuota::new(Some(5));
|
||||
|
||||
assert!(!legacy.check_operation_allowed(4, 2));
|
||||
assert!(durable.uses_durable_reservations());
|
||||
}
|
||||
}
|
||||
|
||||
@@ -13,98 +13,38 @@
|
||||
// limitations under the License.
|
||||
|
||||
pub mod checker;
|
||||
pub(crate) mod reservation;
|
||||
|
||||
use crate::error::Result;
|
||||
use rustfs_config::{
|
||||
QUOTA_API_PATH, QUOTA_EXCEEDED_ERROR_CODE, QUOTA_INTERNAL_ERROR_CODE, QUOTA_INVALID_CONFIG_ERROR_CODE,
|
||||
QUOTA_NOT_FOUND_ERROR_CODE,
|
||||
};
|
||||
use serde::{Deserialize, Deserializer, Serialize, Serializer, de::Error as _};
|
||||
use serde::{Deserialize, Serialize};
|
||||
use thiserror::Error;
|
||||
use time::OffsetDateTime;
|
||||
|
||||
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize, Default)]
|
||||
pub enum QuotaType {
|
||||
/// Hard quota accounting.
|
||||
/// Hard quota: reject immediately when exceeded
|
||||
#[default]
|
||||
#[serde(alias = "HARD", alias = "hard")]
|
||||
Hard,
|
||||
}
|
||||
|
||||
pub(crate) const QUOTA_RESERVATION_PROTOCOL_V1: u32 = 1;
|
||||
|
||||
/// Bucket quota configuration. quota_type defaults to Hard when omitted.
|
||||
#[derive(Debug, Default, Clone, PartialEq)]
|
||||
#[derive(Debug, Deserialize, Serialize, Default, Clone, PartialEq)]
|
||||
pub struct BucketQuota {
|
||||
#[serde(default)]
|
||||
pub quota: Option<u64>,
|
||||
/// Defaults to Hard when missing.
|
||||
#[serde(default)]
|
||||
pub quota_type: QuotaType,
|
||||
/// Optional durable reservation protocol. The wire format gives older
|
||||
/// nodes a zero hard quota so a mixed-version fleet fails closed.
|
||||
pub reservation_protocol: Option<u32>,
|
||||
/// Timestamp when this quota configuration was set (for audit purposes)
|
||||
#[serde(default, with = "time::serde::rfc3339::option")]
|
||||
pub created_at: Option<OffsetDateTime>,
|
||||
/// Accept updated_at for compatibility; not used.
|
||||
pub updated_at: Option<OffsetDateTime>,
|
||||
}
|
||||
|
||||
#[derive(Deserialize, Serialize)]
|
||||
struct BucketQuotaWire {
|
||||
#[serde(default)]
|
||||
quota: Option<u64>,
|
||||
#[serde(default)]
|
||||
quota_type: QuotaType,
|
||||
#[serde(default, skip_serializing_if = "Option::is_none")]
|
||||
reservation_protocol: Option<u32>,
|
||||
#[serde(default, skip_serializing_if = "Option::is_none")]
|
||||
reservation_quota: Option<u64>,
|
||||
#[serde(default, with = "time::serde::rfc3339::option")]
|
||||
created_at: Option<OffsetDateTime>,
|
||||
#[serde(default, with = "time::serde::rfc3339::option", skip_serializing_if = "Option::is_none")]
|
||||
updated_at: Option<OffsetDateTime>,
|
||||
}
|
||||
|
||||
impl Serialize for BucketQuota {
|
||||
fn serialize<S>(&self, serializer: S) -> std::result::Result<S::Ok, S::Error>
|
||||
where
|
||||
S: Serializer,
|
||||
{
|
||||
let durable = self.uses_durable_reservations();
|
||||
BucketQuotaWire {
|
||||
quota: if durable { Some(0) } else { self.quota },
|
||||
quota_type: self.quota_type.clone(),
|
||||
reservation_protocol: self.reservation_protocol,
|
||||
reservation_quota: if durable { self.quota } else { None },
|
||||
created_at: self.created_at,
|
||||
updated_at: self.updated_at,
|
||||
}
|
||||
.serialize(serializer)
|
||||
}
|
||||
}
|
||||
|
||||
impl<'de> Deserialize<'de> for BucketQuota {
|
||||
fn deserialize<D>(deserializer: D) -> std::result::Result<Self, D::Error>
|
||||
where
|
||||
D: Deserializer<'de>,
|
||||
{
|
||||
let wire = BucketQuotaWire::deserialize(deserializer)?;
|
||||
let quota = if wire.reservation_protocol == Some(QUOTA_RESERVATION_PROTOCOL_V1) {
|
||||
Some(
|
||||
wire.reservation_quota
|
||||
.ok_or_else(|| D::Error::custom("reservation_quota is required for reservation protocol v1"))?,
|
||||
)
|
||||
} else {
|
||||
wire.quota
|
||||
};
|
||||
Ok(Self {
|
||||
quota,
|
||||
quota_type: wire.quota_type,
|
||||
reservation_protocol: wire.reservation_protocol,
|
||||
created_at: wire.created_at,
|
||||
updated_at: wire.updated_at,
|
||||
})
|
||||
}
|
||||
pub updated_at: Option<OffsetDateTime>,
|
||||
}
|
||||
|
||||
impl BucketQuota {
|
||||
@@ -123,7 +63,6 @@ impl BucketQuota {
|
||||
Self {
|
||||
quota,
|
||||
quota_type: QuotaType::Hard,
|
||||
reservation_protocol: quota.map(|_| QUOTA_RESERVATION_PROTOCOL_V1),
|
||||
created_at: Some(now),
|
||||
updated_at: None,
|
||||
}
|
||||
@@ -133,19 +72,7 @@ impl BucketQuota {
|
||||
self.quota
|
||||
}
|
||||
|
||||
pub fn uses_durable_reservations(&self) -> bool {
|
||||
self.reservation_protocol == Some(QUOTA_RESERVATION_PROTOCOL_V1)
|
||||
}
|
||||
|
||||
pub fn has_unsupported_reservation_protocol(&self) -> bool {
|
||||
self.reservation_protocol
|
||||
.is_some_and(|version| version != QUOTA_RESERVATION_PROTOCOL_V1)
|
||||
}
|
||||
|
||||
pub fn check_operation_allowed(&self, current_usage: u64, operation_size: u64) -> bool {
|
||||
if operation_size == 0 {
|
||||
return true;
|
||||
}
|
||||
if let Some(quota_limit) = self.quota {
|
||||
current_usage.saturating_add(operation_size) <= quota_limit
|
||||
} else {
|
||||
@@ -167,7 +94,6 @@ pub struct QuotaCheckResult {
|
||||
pub quota_limit: Option<u64>,
|
||||
pub operation_size: u64,
|
||||
pub remaining: Option<u64>,
|
||||
pub uses_durable_reservations: bool,
|
||||
}
|
||||
|
||||
#[derive(Debug)]
|
||||
@@ -193,7 +119,6 @@ pub enum QuotaError {
|
||||
}
|
||||
|
||||
#[derive(Debug, Serialize)]
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub struct QuotaErrorResponse {
|
||||
#[serde(rename = "Code")]
|
||||
pub code: String,
|
||||
@@ -209,7 +134,6 @@ pub struct QuotaErrorResponse {
|
||||
}
|
||||
|
||||
impl QuotaErrorResponse {
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn new(quota_error: &QuotaError, request_id: &str, host_id: &str) -> Self {
|
||||
match quota_error {
|
||||
QuotaError::QuotaExceeded { .. } => Self {
|
||||
@@ -286,59 +210,7 @@ mod tests {
|
||||
let buf = q.marshal_msg().expect("marshal");
|
||||
let restored = BucketQuota::unmarshal(&buf).expect("unmarshal");
|
||||
assert_eq!(q.quota, restored.quota);
|
||||
assert_eq!(restored.quota_type, QuotaType::Hard);
|
||||
assert_eq!(restored.reservation_protocol, Some(QUOTA_RESERVATION_PROTOCOL_V1));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn clearing_quota_keeps_the_legacy_compatible_type() {
|
||||
let quota = BucketQuota::new(None);
|
||||
|
||||
assert_eq!(quota.quota_type, QuotaType::Hard);
|
||||
assert_eq!(quota.reservation_protocol, None);
|
||||
assert!(!quota.uses_durable_reservations());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn durable_quota_makes_legacy_nodes_fail_closed() {
|
||||
let json = serde_json::to_vec(&BucketQuota::new(Some(2048))).expect("durable quota should serialize");
|
||||
let quota: BucketQuota = serde_json::from_slice(&json).expect("current quota version should parse");
|
||||
assert!(quota.uses_durable_reservations());
|
||||
assert_eq!(quota.quota, Some(2048));
|
||||
|
||||
#[derive(Deserialize)]
|
||||
enum LegacyQuotaType {
|
||||
Hard,
|
||||
}
|
||||
#[derive(Deserialize)]
|
||||
struct LegacyBucketQuota {
|
||||
#[allow(dead_code)]
|
||||
quota: Option<u64>,
|
||||
#[allow(dead_code)]
|
||||
quota_type: LegacyQuotaType,
|
||||
}
|
||||
let legacy = serde_json::from_slice::<LegacyBucketQuota>(&json)
|
||||
.expect("legacy readers should ignore the reservation protocol field");
|
||||
assert_eq!(legacy.quota, Some(0));
|
||||
assert!(matches!(legacy.quota_type, LegacyQuotaType::Hard));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn unknown_reservation_protocol_does_not_activate_v1() {
|
||||
let quota: BucketQuota =
|
||||
serde_json::from_str(r#"{"quota":0,"quota_type":"Hard","reservation_protocol":2,"reservation_quota":2048}"#)
|
||||
.expect("future protocol should remain parseable");
|
||||
|
||||
assert!(!quota.uses_durable_reservations());
|
||||
assert!(quota.has_unsupported_reservation_protocol());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn reservation_protocol_v1_requires_reservation_quota() {
|
||||
let err = serde_json::from_str::<BucketQuota>(r#"{"quota":0,"quota_type":"Hard","reservation_protocol":1}"#)
|
||||
.expect_err("v1 without its authoritative quota must fail closed");
|
||||
|
||||
assert!(err.to_string().contains("reservation_quota is required"));
|
||||
assert_eq!(q.quota_type, restored.quota_type);
|
||||
}
|
||||
|
||||
/// unmarshal accepts format without quota_type
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -60,7 +60,7 @@ pub use replication_filemeta_boundary::{
|
||||
pub(crate) use replication_filemeta_boundary::{
|
||||
replication_state_from_filemeta, replication_status_from_filemeta, version_purge_status_from_filemeta,
|
||||
};
|
||||
pub(crate) use replication_lifecycle_bridge::ReplicationLifecycleBridge;
|
||||
pub(crate) use replication_lifecycle_bridge::{ReplicationLifecycleBridge, ReplicationLifecycleConfig};
|
||||
pub(crate) use replication_migration_bridge::ReplicationMigrationBridge;
|
||||
pub use replication_object_bridge::ReplicationObjectBridge;
|
||||
pub use replication_object_config::{DeleteReplicationConfigSnapshot, ReplicationConfig};
|
||||
@@ -81,6 +81,6 @@ pub use replication_queue_boundary::{
|
||||
pub use replication_resync_boundary::{BucketReplicationResyncStatus, ResyncOpts, TargetReplicationResyncStatus};
|
||||
pub use replication_scanner_bridge::ReplicationScannerBridge;
|
||||
pub use replication_state::{ReplicationStats, RuntimeReplicationTargetBacklog};
|
||||
pub use replication_stats_boundary::{BucketReplicationStat, BucketReplicationStats, BucketStats, InQueueMetric, XferStats};
|
||||
pub use replication_stats_boundary::{BucketReplicationStats, BucketStats};
|
||||
pub use replication_storage_boundary::{ReplicationObjectIO, ReplicationStorage};
|
||||
pub(crate) use replication_target_config_bridge::ReplicationTargetConfigBridge;
|
||||
|
||||
@@ -37,10 +37,6 @@ impl ReplicationConfigStore {
|
||||
com::read_config_limited(api, file, max_bytes).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) async fn read_no_lock<S>(api: Arc<S>, file: &str) -> Result<Vec<u8>>
|
||||
where
|
||||
S: ReplicationObjectIO,
|
||||
|
||||
@@ -24,27 +24,15 @@ use super::replication_storage_boundary::{
|
||||
DeletedObject, ObjectInfo, ObjectOptions, ObjectToDelete, deleted_object_for_replication,
|
||||
};
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) type ReplicationLifecycleConfig = ReplicationConfig;
|
||||
|
||||
pub(crate) struct ReplicationLifecycleBridge;
|
||||
|
||||
impl ReplicationLifecycleBridge {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn new_config(config: ReplicationConfiguration) -> ReplicationLifecycleConfig {
|
||||
ReplicationConfig::new(Some(config), None)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn has_pending_version_purge(
|
||||
config: &ReplicationLifecycleConfig,
|
||||
object_name: &str,
|
||||
@@ -57,10 +45,6 @@ impl ReplicationLifecycleBridge {
|
||||
.is_some_and(|config| config.has_active_rules(object_name, true))
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) async fn check_delete_replication(
|
||||
bucket: &str,
|
||||
object: &ObjectToDelete,
|
||||
@@ -70,10 +54,6 @@ impl ReplicationLifecycleBridge {
|
||||
check_replicate_delete(bucket, object, source, opts, None).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn version_delete_replication_state(decision: &ReplicateDecision) -> ReplicationState {
|
||||
let pending_status = decision.pending_status();
|
||||
ReplicationState {
|
||||
|
||||
@@ -19,33 +19,17 @@ use time::OffsetDateTime;
|
||||
use super::replication_error_boundary::Result;
|
||||
use crate::bucket::msgp_decode;
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) struct ReplicationMsgpCodec;
|
||||
|
||||
impl ReplicationMsgpCodec {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn read_ext8_time<R: Read>(rd: &mut R) -> Result<OffsetDateTime> {
|
||||
msgp_decode::read_msgp_ext8_time(rd)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn skip_value<R: Read>(rd: &mut R) -> Result<()> {
|
||||
msgp_decode::skip_msgp_value(rd)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn write_time<W: Write>(wr: &mut W, time: OffsetDateTime) -> Result<()> {
|
||||
msgp_decode::write_msgp_time(wr, time)
|
||||
}
|
||||
|
||||
@@ -77,10 +77,6 @@ impl ReplicationObjectBridge {
|
||||
load_delete_request_config_in(ctx, bucket).await
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) async fn delete_config_snapshot_in(
|
||||
ctx: &ReplicationInstanceContext,
|
||||
bucket: &str,
|
||||
|
||||
@@ -231,10 +231,6 @@ pub(crate) async fn load_delete_replication_config(
|
||||
delete_snapshot_from_metadata(ReplicationMetadataStore::delete_metadata(bucket).await?)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) async fn load_delete_replication_config_in(
|
||||
ctx: &ReplicationInstanceContext,
|
||||
bucket: &str,
|
||||
|
||||
@@ -217,10 +217,6 @@ impl DurableMrfBacklogTracker {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn durable_mrf_backlog_tracker_from_entries(entries: &[MrfReplicateEntry]) -> DurableMrfBacklogTracker {
|
||||
let mut tracker = DurableMrfBacklogTracker {
|
||||
available: true,
|
||||
@@ -716,10 +712,6 @@ pub struct ReplicationPool<S: ReplicationStorage> {
|
||||
|
||||
// MRF worker lifecycle
|
||||
mrf_worker_cancellations: Mutex<Vec<CancellationToken>>,
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
mrf_stop_tx: Sender<()>,
|
||||
|
||||
// Worker size tracking
|
||||
@@ -948,10 +940,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Resizes worker priority and counts
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub async fn resize_worker_priority(
|
||||
&self,
|
||||
pri: ReplicationPriority,
|
||||
@@ -1192,10 +1180,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Queues an MRF save operation
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
async fn queue_mrf_save(&self, entry: MrfReplicateEntry) {
|
||||
let _ = self.queue_mrf_save_admission(entry, "mrf_worker").await;
|
||||
}
|
||||
@@ -1667,10 +1651,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Worker function for handling regular replication operations
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
async fn add_worker(
|
||||
&self,
|
||||
mut rx: Receiver<ReplicationOperation>,
|
||||
@@ -1684,10 +1664,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Worker function for handling large object replication operations
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
async fn add_large_worker(
|
||||
&self,
|
||||
mut rx: Receiver<ReplicationOperation>,
|
||||
@@ -1702,10 +1678,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Worker function for handling MRF (Most Recent Failures) operations
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
async fn add_mrf_worker(
|
||||
&self,
|
||||
mut rx: Receiver<ReplicationOperation>,
|
||||
@@ -1719,10 +1691,6 @@ impl<S: ReplicationStorage> ReplicationPool<S> {
|
||||
}
|
||||
|
||||
/// Delete resync metadata from replication resync state in memory
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub async fn delete_resync_metadata(&self, bucket: &str) {
|
||||
let mut status_map = self.resyncer.status_map.write().await;
|
||||
status_map.remove(bucket);
|
||||
|
||||
@@ -21,31 +21,11 @@ pub(crate) use rustfs_replication::{
|
||||
should_count_head_proxy_failure,
|
||||
};
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const RESYNC_META_FORMAT: u16 = rustfs_replication::resync::RESYNC_META_FORMAT;
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const RESYNC_META_VERSION: u16 = rustfs_replication::resync::RESYNC_META_VERSION;
|
||||
pub(crate) const RESYNC_FILE_MAX_BYTES: usize = rustfs_replication::RESYNC_FILE_MAX_BYTES;
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const WIRE_ZERO_TIME_UNIX: i64 = rustfs_replication::resync::WIRE_ZERO_TIME_UNIX;
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const MRF_META_FORMAT: u16 = rustfs_replication::mrf::MRF_META_FORMAT;
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "declared boundary surface for the ECStore replication split plan; no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const MRF_META_VERSION: u16 = rustfs_replication::mrf::MRF_META_VERSION;
|
||||
|
||||
fn map_replication_error(err: rustfs_replication::Error) -> Error {
|
||||
|
||||
@@ -122,10 +122,6 @@ const REPLICATION_TARGET_OFFLINE_ERROR_MARKERS: &[&str] = &[
|
||||
"tcp connect error",
|
||||
];
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
const RESYNC_TIME_INTERVAL: TokioDuration = TokioDuration::from_secs(60);
|
||||
|
||||
static WARNED_MONITOR_UNINIT: std::sync::Once = std::sync::Once::new();
|
||||
@@ -332,10 +328,6 @@ fn bounded_resync_max_jobs(value: usize) -> usize {
|
||||
#[derive(Debug)]
|
||||
pub struct ReplicationResyncer {
|
||||
pub status_map: Arc<RwLock<HashMap<String, BucketReplicationResyncStatus>>>,
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub worker_size: usize,
|
||||
pub(crate) cancel_tokens: Arc<RwLock<HashMap<ResyncCancelKey, CancellationToken>>>,
|
||||
resync_admission: Arc<Semaphore>,
|
||||
@@ -552,10 +544,6 @@ impl ReplicationResyncer {
|
||||
.is_some_and(|status| status.failed_count > 0)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub async fn persist_to_disk<S>(&self, cancel_token: CancellationToken, api: Arc<S>)
|
||||
where
|
||||
S: ReplicationObjectIO,
|
||||
|
||||
@@ -340,10 +340,6 @@ impl ReplicationStats {
|
||||
}
|
||||
|
||||
/// Site replication update replica statistics
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity replication surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn sr_update_replica_stat(&self, size: i64) {
|
||||
self.sr_stats.replica_size.fetch_add(size, Ordering::Relaxed);
|
||||
self.sr_stats.replica_count.fetch_add(1, Ordering::Relaxed);
|
||||
@@ -708,12 +704,6 @@ impl ReplicationStats {
|
||||
} else {
|
||||
BucketReplicationStats::new()
|
||||
};
|
||||
// Stamp the serializable failure windows from the live samples: the
|
||||
// samples themselves do not cross the peer-RPC wire, so this snapshot
|
||||
// is what cluster aggregation and the metrics endpoints see.
|
||||
for stat in replication_stats.stats.values_mut() {
|
||||
stat.fail_stats.refresh_windows();
|
||||
}
|
||||
let uptime = if cache.contains_key(bucket) {
|
||||
SystemTime::now()
|
||||
.duration_since(SystemTime::UNIX_EPOCH)
|
||||
|
||||
@@ -15,9 +15,7 @@
|
||||
#[cfg(test)]
|
||||
pub(crate) use rustfs_replication::FailStats;
|
||||
pub(crate) use rustfs_replication::{
|
||||
ActiveWorkerStat, ProxyMetric, ProxyStatsCache, QueueCache, ReplicationMetricScope, SRMetricsSummary,
|
||||
ActiveWorkerStat, BucketReplicationStat, InQueueMetric, ProxyMetric, ProxyStatsCache, QueueCache, ReplicationMetricScope,
|
||||
SRMetricsSummary, XferStats,
|
||||
};
|
||||
// Public so the admin wire DTOs (rustfs/src/admin/replication_metrics_wire.rs)
|
||||
// can project the internal stats onto the minio-go response shapes through
|
||||
// the storage_api facade chain.
|
||||
pub use rustfs_replication::{BucketReplicationStat, BucketReplicationStats, BucketStats, InQueueMetric, XferStats};
|
||||
pub use rustfs_replication::{BucketReplicationStats, BucketStats};
|
||||
|
||||
@@ -27,10 +27,8 @@ use rustfs_utils::http::{
|
||||
AMZ_OBJECT_TAGGING, AMZ_SERVER_SIDE_ENCRYPTION, AMZ_SERVER_SIDE_ENCRYPTION_KMS_CONTEXT, AMZ_SERVER_SIDE_ENCRYPTION_KMS_ID,
|
||||
AMZ_STORAGE_CLASS, AMZ_TAG_COUNT, CACHE_CONTROL, CONTENT_DISPOSITION, CONTENT_ENCODING, CONTENT_LANGUAGE, CONTENT_TYPE,
|
||||
HeaderExt as _, SUFFIX_OBJECTLOCK_LEGALHOLD_TIMESTAMP, SUFFIX_OBJECTLOCK_RETENTION_TIMESTAMP,
|
||||
SUFFIX_REPLICATION_ACTUAL_OBJECT_SIZE, SUFFIX_REPLICATION_SSEC_CRC, SUFFIX_SOURCE_REPLICATION_LEGALHOLD_TIMESTAMP,
|
||||
SUFFIX_SOURCE_REPLICATION_RETENTION_TIMESTAMP, SUFFIX_SOURCE_REPLICATION_TAGGING_TIMESTAMP, SUFFIX_TAGGING_TIMESTAMP,
|
||||
get_str, insert_header_map, is_internal_key, is_object_encryption_marker, is_replication_stripped_encryption_key,
|
||||
ssec_replication_transport_header,
|
||||
SUFFIX_REPLICATION_ACTUAL_OBJECT_SIZE, SUFFIX_REPLICATION_SSEC_CRC, SUFFIX_TAGGING_TIMESTAMP, get_str, insert_header_map,
|
||||
is_internal_key, is_object_encryption_marker, is_replication_stripped_encryption_key, ssec_replication_transport_header,
|
||||
};
|
||||
use time::OffsetDateTime;
|
||||
use time::format_description::well_known::Rfc3339;
|
||||
@@ -121,27 +119,6 @@ fn classify_replication_source_encryption(metadata: &HashMap<String, String>) ->
|
||||
}
|
||||
}
|
||||
|
||||
fn is_legacy_source_replication_timestamp_key(key: &str) -> bool {
|
||||
fn has_prefix_and_suffix(key: &str, prefix: &str, suffix: &str) -> bool {
|
||||
let key = key.as_bytes();
|
||||
key.len() == prefix.len() + suffix.len()
|
||||
&& key[..prefix.len()].eq_ignore_ascii_case(prefix.as_bytes())
|
||||
&& key[prefix.len()..].eq_ignore_ascii_case(suffix.as_bytes())
|
||||
}
|
||||
|
||||
[
|
||||
SUFFIX_SOURCE_REPLICATION_TAGGING_TIMESTAMP,
|
||||
SUFFIX_SOURCE_REPLICATION_RETENTION_TIMESTAMP,
|
||||
SUFFIX_SOURCE_REPLICATION_LEGALHOLD_TIMESTAMP,
|
||||
]
|
||||
.iter()
|
||||
.any(|suffix| {
|
||||
["x-rustfs-", "x-minio-"]
|
||||
.iter()
|
||||
.any(|prefix| has_prefix_and_suffix(key, prefix, suffix))
|
||||
})
|
||||
}
|
||||
|
||||
pub(crate) fn replication_object_is_ssec_encrypted(user_defined: &HashMap<String, String>) -> bool {
|
||||
rustfs_replication::is_ssec_encrypted(user_defined)
|
||||
}
|
||||
@@ -199,11 +176,6 @@ pub(crate) fn replication_put_object_options(sc: &str, object_info: &ObjectInfo)
|
||||
continue;
|
||||
}
|
||||
|
||||
if is_legacy_source_replication_timestamp_key(key) {
|
||||
meta.insert(format!("x-amz-meta-{key}"), value.to_string());
|
||||
continue;
|
||||
}
|
||||
|
||||
if is_internal_key(key) || is_standard_header(key) {
|
||||
continue;
|
||||
}
|
||||
@@ -287,23 +259,15 @@ pub(crate) fn replication_put_object_options(sc: &str, object_info: &ObjectInfo)
|
||||
|
||||
if !tags.is_empty() {
|
||||
put_options.user_tags = tags;
|
||||
put_options.internal.tagging_timestamp =
|
||||
if let Some(timestamp) = get_str(&object_info.user_defined, SUFFIX_TAGGING_TIMESTAMP) {
|
||||
OffsetDateTime::parse(×tamp, &Rfc3339)
|
||||
.map_err(|err| Error::other(format!("Failed to parse tagging timestamp: {err}")))?
|
||||
} else {
|
||||
object_info.mod_time.unwrap_or(OffsetDateTime::UNIX_EPOCH)
|
||||
};
|
||||
}
|
||||
}
|
||||
// Load the stored tagging timestamp independently of whether any tags
|
||||
// remain: DeleteObjectTagging leaves the object tagless but stamps this
|
||||
// key, and the deletion's LWW timestamp must still reach the replica.
|
||||
// With no stored key, fall back to mod_time only while tags exist
|
||||
// (MinIO parity); a tagless object without the key was never tagged and
|
||||
// keeps the epoch default (no header).
|
||||
put_options.internal.tagging_timestamp = if let Some(timestamp) = get_str(&object_info.user_defined, SUFFIX_TAGGING_TIMESTAMP)
|
||||
{
|
||||
OffsetDateTime::parse(×tamp, &Rfc3339)
|
||||
.map_err(|err| Error::other(format!("Failed to parse tagging timestamp: {err}")))?
|
||||
} else if !put_options.user_tags.is_empty() {
|
||||
object_info.mod_time.unwrap_or(OffsetDateTime::UNIX_EPOCH)
|
||||
} else {
|
||||
OffsetDateTime::UNIX_EPOCH
|
||||
};
|
||||
|
||||
let metadata = &*object_info.user_defined;
|
||||
|
||||
@@ -319,15 +283,13 @@ pub(crate) fn replication_put_object_options(sc: &str, object_info: &ObjectInfo)
|
||||
put_options.cache_control = cache_control.to_string();
|
||||
}
|
||||
|
||||
if let Some(mode) = metadata.lookup(AMZ_OBJECT_LOCK_MODE).filter(|mode| !mode.is_empty()) {
|
||||
if let Some(mode) = metadata.lookup(AMZ_OBJECT_LOCK_MODE) {
|
||||
put_options.mode = Some(ObjectLockRetentionMode::from(mode.to_uppercase().as_str()));
|
||||
}
|
||||
|
||||
if let Some(retain_until_date) = metadata.lookup(AMZ_OBJECT_LOCK_RETAIN_UNTIL_DATE) {
|
||||
if !retain_until_date.is_empty() {
|
||||
put_options.retain_until_date = OffsetDateTime::parse(retain_until_date, &Rfc3339)
|
||||
.map_err(|err| Error::other(format!("Failed to parse retain until date: {err}")))?;
|
||||
}
|
||||
put_options.retain_until_date = OffsetDateTime::parse(retain_until_date, &Rfc3339)
|
||||
.map_err(|err| Error::other(format!("Failed to parse retain until date: {err}")))?;
|
||||
put_options.internal.retention_timestamp =
|
||||
if let Some(timestamp) = get_str(&object_info.user_defined, SUFFIX_OBJECTLOCK_RETENTION_TIMESTAMP) {
|
||||
OffsetDateTime::parse(×tamp, &Rfc3339).unwrap_or(OffsetDateTime::UNIX_EPOCH)
|
||||
@@ -732,110 +694,6 @@ mod tests {
|
||||
assert!(options.internal.replication_request);
|
||||
}
|
||||
|
||||
/// DeleteObjectTagging leaves the object tagless but stamps the
|
||||
/// tagging-timestamp internal key; the deletion's LWW timestamp must
|
||||
/// still be loaded (and therefore sent) so the replica can order the
|
||||
/// deletion against concurrent tag edits.
|
||||
#[test]
|
||||
fn replication_put_options_carry_tagging_timestamp_after_tag_deletion() {
|
||||
let mut metadata = std::collections::HashMap::new();
|
||||
rustfs_utils::http::insert_str(&mut metadata, SUFFIX_TAGGING_TIMESTAMP, "2026-01-02T03:04:05Z".to_string());
|
||||
|
||||
let object_info = ObjectInfo {
|
||||
user_defined: Arc::new(metadata),
|
||||
user_tags: Arc::new(String::new()),
|
||||
mod_time: Some(OffsetDateTime::UNIX_EPOCH),
|
||||
version_id: Some(Uuid::nil()),
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
let (options, _) = replication_put_object_options("", &object_info).expect("build put options");
|
||||
|
||||
assert!(options.user_tags.is_empty());
|
||||
assert_eq!(
|
||||
options.internal.tagging_timestamp,
|
||||
OffsetDateTime::parse("2026-01-02T03:04:05Z", &Rfc3339).expect("valid timestamp"),
|
||||
"the stored tagging timestamp must load independently of remaining tags"
|
||||
);
|
||||
|
||||
// A tagless object without the stored key was never tagged: the epoch
|
||||
// default keeps the header unsent.
|
||||
let untagged = ObjectInfo {
|
||||
user_tags: Arc::new(String::new()),
|
||||
mod_time: Some(OffsetDateTime::from_unix_timestamp(1_700_000_000).expect("timestamp")),
|
||||
version_id: Some(Uuid::nil()),
|
||||
..Default::default()
|
||||
};
|
||||
let (options, _) = replication_put_object_options("", &untagged).expect("build put options");
|
||||
assert_eq!(options.internal.tagging_timestamp, OffsetDateTime::UNIX_EPOCH);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn replication_put_options_do_not_promote_legacy_user_timestamp_metadata() {
|
||||
let legacy_keys = [
|
||||
"x-rustfs-source-replication-tagging-timestamp",
|
||||
"x-rustfs-source-replication-retention-timestamp",
|
||||
"x-rustfs-source-replication-legalhold-timestamp",
|
||||
"x-minio-source-replication-tagging-timestamp",
|
||||
"x-minio-source-replication-retention-timestamp",
|
||||
"x-minio-source-replication-legalhold-timestamp",
|
||||
];
|
||||
let object_info = ObjectInfo {
|
||||
user_defined: Arc::new(
|
||||
legacy_keys
|
||||
.iter()
|
||||
.map(|key| (key.to_string(), "2099-01-02T03:04:05Z".to_string()))
|
||||
.collect(),
|
||||
),
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
let (options, _) = replication_put_object_options("", &object_info).expect("build put options");
|
||||
|
||||
for legacy_key in legacy_keys {
|
||||
assert!(!options.user_metadata.contains_key(legacy_key));
|
||||
assert_eq!(
|
||||
options
|
||||
.user_metadata
|
||||
.get(&format!("x-amz-meta-{legacy_key}"))
|
||||
.map(String::as_str),
|
||||
Some("2099-01-02T03:04:05Z")
|
||||
);
|
||||
}
|
||||
assert_eq!(options.internal.tagging_timestamp, OffsetDateTime::UNIX_EPOCH);
|
||||
assert_eq!(options.internal.retention_timestamp, OffsetDateTime::UNIX_EPOCH);
|
||||
assert_eq!(options.internal.legalhold_timestamp, OffsetDateTime::UNIX_EPOCH);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn replication_put_options_carry_retention_timestamp_after_clear() {
|
||||
let mut metadata = HashMap::from([
|
||||
(AMZ_OBJECT_LOCK_MODE.to_string(), String::new()),
|
||||
(AMZ_OBJECT_LOCK_RETAIN_UNTIL_DATE.to_string(), String::new()),
|
||||
]);
|
||||
rustfs_utils::http::insert_str(&mut metadata, SUFFIX_OBJECTLOCK_RETENTION_TIMESTAMP, "2026-01-02T03:04:05Z".to_string());
|
||||
let object_info = ObjectInfo {
|
||||
user_defined: Arc::new(metadata),
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
let (options, _) = replication_put_object_options("", &object_info).expect("retention clear must replicate");
|
||||
|
||||
assert!(options.mode.is_none());
|
||||
assert_eq!(options.retain_until_date, OffsetDateTime::UNIX_EPOCH);
|
||||
assert_eq!(
|
||||
options.internal.retention_timestamp,
|
||||
OffsetDateTime::parse("2026-01-02T03:04:05Z", &Rfc3339).expect("valid timestamp")
|
||||
);
|
||||
let headers = options.header();
|
||||
assert!(!headers.contains_key(AMZ_OBJECT_LOCK_MODE));
|
||||
assert!(!headers.contains_key(AMZ_OBJECT_LOCK_RETAIN_UNTIL_DATE));
|
||||
assert_eq!(
|
||||
rustfs_utils::http::get_header(&headers, SUFFIX_SOURCE_REPLICATION_RETENTION_TIMESTAMP).as_deref(),
|
||||
Some("2026-01-02T03:04:05Z")
|
||||
);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn replication_put_options_strip_encryption_metadata_from_plaintext_objects() {
|
||||
use rustfs_utils::http::object_encryption_keys::{INTERNAL_ENCRYPTION_ORIGINAL_SIZE_HEADER, SSEC_ORIGINAL_SIZE_HEADER};
|
||||
|
||||
@@ -40,14 +40,7 @@ impl ARN {
|
||||
|
||||
impl Display for ARN {
|
||||
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
||||
// The `minio` partition is deliberate: madmin-go's ParseARN
|
||||
// hard-rejects any other partition, so native mc/madmin tooling can
|
||||
// only decode remote-target ARNs minted in this form (backlog#1675
|
||||
// P1-7). Legacy `arn:rustfs:` ARNs persisted by older releases stay
|
||||
// readable via the FromStr whitelist below; runtime matching between
|
||||
// targets and replication rules is by full-string equality, so mixed
|
||||
// partitions coexist safely.
|
||||
write!(f, "arn:minio:{}:{}:{}:{}", self.arn_type, self.region, self.id, self.bucket)
|
||||
write!(f, "arn:rustfs:{}:{}:{}:{}", self.arn_type, self.region, self.id, self.bucket)
|
||||
}
|
||||
}
|
||||
|
||||
@@ -55,12 +48,7 @@ impl FromStr for ARN {
|
||||
type Err = std::io::Error;
|
||||
|
||||
fn from_str(s: &str) -> Result<Self, Self::Err> {
|
||||
// Partition whitelist, not just an `arn:` check: `BucketTargetType::
|
||||
// from_str(...).unwrap_or_default()` below never fails, so this is
|
||||
// the only structural gate rejecting foreign ARNs. `arn:rustfs:` is
|
||||
// the legacy partition and must stay accepted forever (persisted
|
||||
// bucket-targets.json / replication configs from older releases).
|
||||
if !s.starts_with("arn:minio:") && !s.starts_with("arn:rustfs:") {
|
||||
if !s.starts_with("arn:rustfs:") {
|
||||
return Err(std::io::Error::new(std::io::ErrorKind::InvalidInput, "Invalid ARN format"));
|
||||
}
|
||||
|
||||
@@ -113,50 +101,14 @@ mod tests {
|
||||
}
|
||||
|
||||
/// RustFS commonly generates ARNs with an empty region:
|
||||
/// `arn:minio:replication::<deployment_id>:<bucket>`.
|
||||
/// `arn:rustfs:replication::<deployment_id>:<bucket>`.
|
||||
#[test]
|
||||
fn from_str_handles_empty_region_segment() {
|
||||
let parsed = ARN::from_str("arn:minio:replication::depl-123:bucket-a").expect("valid ARN must parse");
|
||||
let parsed = ARN::from_str("arn:rustfs:replication::depl-123:bucket-a").expect("valid ARN must parse");
|
||||
|
||||
assert_eq!(parsed.arn_type, BucketTargetType::ReplicationService);
|
||||
assert_eq!(parsed.region, "", "region segment is empty in this form");
|
||||
assert_eq!(parsed.id, "depl-123");
|
||||
assert_eq!(parsed.bucket, "bucket-a");
|
||||
}
|
||||
|
||||
/// madmin-go's `ParseARN` hard-rejects anything that does not start with
|
||||
/// `arn:minio:`, so generated ARNs must use the `minio` partition or the
|
||||
/// native mc/madmin tooling cannot decode remote-target listings.
|
||||
#[test]
|
||||
fn display_emits_minio_partition() {
|
||||
let arn = ARN::new(
|
||||
BucketTargetType::ReplicationService,
|
||||
"depl-123".to_string(),
|
||||
String::new(),
|
||||
"bucket-a".to_string(),
|
||||
);
|
||||
|
||||
assert_eq!(arn.to_string(), "arn:minio:replication::depl-123:bucket-a");
|
||||
}
|
||||
|
||||
/// Persisted bucket-targets.json files from older RustFS releases carry
|
||||
/// `arn:rustfs:` ARNs; the legacy partition must stay parseable forever.
|
||||
#[test]
|
||||
fn from_str_accepts_legacy_rustfs_partition() {
|
||||
let parsed = ARN::from_str("arn:rustfs:replication:us-east-1:depl-123:bucket-a").expect("legacy ARN must parse");
|
||||
|
||||
assert_eq!(parsed.arn_type, BucketTargetType::ReplicationService);
|
||||
assert_eq!(parsed.region, "us-east-1");
|
||||
assert_eq!(parsed.id, "depl-123");
|
||||
assert_eq!(parsed.bucket, "bucket-a");
|
||||
}
|
||||
|
||||
/// The partition whitelist is the only structural gate: `BucketTargetType::
|
||||
/// from_str(...).unwrap_or_default()` never fails, so any 6-segment string
|
||||
/// would otherwise parse as `type=None`.
|
||||
#[test]
|
||||
fn from_str_rejects_unknown_partition() {
|
||||
assert!(ARN::from_str("arn:aws:replication::depl-123:bucket-a").is_err());
|
||||
assert!(ARN::from_str("not-an-arn").is_err());
|
||||
}
|
||||
}
|
||||
|
||||
@@ -59,10 +59,6 @@ impl fmt::Debug for Credentials {
|
||||
}
|
||||
|
||||
#[derive(Debug, Deserialize, Serialize, Default, Clone)]
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity bucket-target service discriminator with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub enum ServiceType {
|
||||
#[default]
|
||||
Replication,
|
||||
|
||||
@@ -73,6 +73,23 @@ pub fn check_valid_bucket_name_strict(bucket_name: &str) -> Result<()> {
|
||||
check_bucket_name_common(bucket_name, true)
|
||||
}
|
||||
|
||||
pub fn check_valid_object_name_prefix(object_name: &str) -> Result<()> {
|
||||
if object_name.len() > 1024 {
|
||||
return Err(Error::other("Object name cannot be longer than 1024 characters"));
|
||||
}
|
||||
if !object_name.is_ascii() {
|
||||
return Err(Error::other("Object name with non-UTF-8 strings are not supported"));
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
pub fn check_valid_object_name(object_name: &str) -> Result<()> {
|
||||
if object_name.trim().is_empty() {
|
||||
return Err(Error::other("Object name cannot be empty"));
|
||||
}
|
||||
check_valid_object_name_prefix(object_name)
|
||||
}
|
||||
|
||||
pub fn deserialize<T>(input: &[u8]) -> xml::DeResult<T>
|
||||
where
|
||||
T: for<'xml> xml::Deserialize<'xml>,
|
||||
@@ -83,10 +100,6 @@ where
|
||||
Ok(ans)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "xml serialize helper with no caller in this port; the live sibling is deserialize (backlog#1823)"
|
||||
)]
|
||||
pub fn serialize_content<T: xml::SerializeContent>(val: &T) -> xml::SerResult<String> {
|
||||
let mut buf = Vec::with_capacity(256);
|
||||
{
|
||||
@@ -173,27 +186,15 @@ pub fn is_valid_object_name(object: &str) -> bool {
|
||||
/// Client-facing reason attached to rejections of object keys that Win32/NTFS
|
||||
/// cannot represent as file paths (issue #3299). Deployments on Linux/macOS
|
||||
/// accept the full S3 key character set.
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live on Windows: callers sit inside the #[cfg(target_os = \"windows\")] block in check_object_name_for_length_and_slash (backlog#1823)"
|
||||
)]
|
||||
pub const WINDOWS_RESERVED_CHARACTERS_REASON: &str =
|
||||
"object key contains characters unsupported on Windows hosts (one of ':', '*', '?', '\"', '|', '<', '>')";
|
||||
|
||||
/// Client-facing reason for path segments Windows can store but not address
|
||||
/// afterwards (issue #3449): trailing dot/space or reserved DOS device names.
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live on Windows: callers sit inside the #[cfg(target_os = \"windows\")] block in check_object_name_for_length_and_slash (backlog#1823)"
|
||||
)]
|
||||
pub const WINDOWS_RESERVED_SEGMENT_REASON: &str = "object key contains a path segment unsupported on Windows hosts (trailing dot or space, or a reserved device name such as NUL/CON/COM1)";
|
||||
|
||||
/// Reserved DOS device names that shadow regular files on Windows, even when
|
||||
/// an extension is appended (e.g. `NUL.txt` resolves to the `NUL` device).
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live on Windows: callers sit inside the #[cfg(target_os = \"windows\")] block in check_object_name_for_length_and_slash (backlog#1823)"
|
||||
)]
|
||||
const WINDOWS_RESERVED_NAMES: &[&str] = &[
|
||||
"CON", "PRN", "AUX", "NUL", "COM1", "COM2", "COM3", "COM4", "COM5", "COM6", "COM7", "COM8", "COM9", "LPT1", "LPT2", "LPT3",
|
||||
"LPT4", "LPT5", "LPT6", "LPT7", "LPT8", "LPT9",
|
||||
@@ -203,10 +204,6 @@ const WINDOWS_RESERVED_NAMES: &[&str] = &[
|
||||
/// the Win32 API cannot address afterwards (issue #3449): segments ending in a
|
||||
/// dot or a space, and reserved DOS device names — bare or with an extension
|
||||
/// (`NUL.txt`), matching classic Win32 path resolution semantics.
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live on Windows: callers sit inside the #[cfg(target_os = \"windows\")] block in check_object_name_for_length_and_slash (backlog#1823)"
|
||||
)]
|
||||
pub fn object_name_has_windows_incompatible_segment(object: &str) -> bool {
|
||||
object.split(['/', '\\']).any(|segment| {
|
||||
if segment.ends_with('.') || segment.ends_with(' ') {
|
||||
|
||||
@@ -90,10 +90,6 @@ impl BucketVersioningSys {
|
||||
/// caller's own instance context so a second in-process store never
|
||||
/// answers with the first instance's versioning state; falls back to the
|
||||
/// ambient system when the instance cell is not initialized.
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "instance-scoped seam (backlog#1052) with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
pub(crate) async fn get_in(ctx: &crate::runtime::instance::InstanceContext, bucket: &str) -> Result<VersioningConfiguration> {
|
||||
if bucket == RUSTFS_META_BUCKET || bucket.starts_with(RUSTFS_META_BUCKET) {
|
||||
return Ok(VersioningConfiguration::default());
|
||||
|
||||
@@ -15,7 +15,6 @@
|
||||
use crate::disk::disk_store::{get_drive_walkdir_peek_timeout, get_drive_walkdir_stall_timeout};
|
||||
use crate::disk::error::DiskError;
|
||||
use crate::disk::{self, DiskAPI, DiskStore, WalkDirOptions};
|
||||
use futures::future::join_all;
|
||||
use metrics::counter;
|
||||
use rustfs_filemeta::{MetaCacheEntries, MetaCacheEntry, MetacacheReader, is_io_eof};
|
||||
use std::{
|
||||
@@ -656,7 +655,6 @@ async fn list_path_raw_inner(
|
||||
errs.push(None);
|
||||
}
|
||||
let mut pending_entries: Vec<Option<MetaCacheEntry>> = vec![None; readers.len()];
|
||||
let mut peek_outcomes: Vec<Option<PeekOutcome>> = std::iter::repeat_with(|| None).take(readers.len()).collect();
|
||||
|
||||
loop {
|
||||
let mut current = MetaCacheEntry::default();
|
||||
@@ -678,21 +676,6 @@ async fn list_path_raw_inner(
|
||||
let mut has_err = 0;
|
||||
let mut agree = 0;
|
||||
|
||||
// Start every missing head read in the same round so one stalled
|
||||
// disk cannot multiply the wait budget by the erasure-set width.
|
||||
// Outcomes are still consumed below in stable disk-index order.
|
||||
let concurrent_peeks = readers.iter_mut().enumerate().filter_map(|(i, reader)| {
|
||||
if errs[i].is_some() || pending_entries[i].is_some() {
|
||||
return None;
|
||||
}
|
||||
|
||||
let cancel = &revjob_rx;
|
||||
Some(async move { (i, peek_with_timeout(cancel, reader, peek_timeout).await) })
|
||||
});
|
||||
for (i, outcome) in join_all(concurrent_peeks).await {
|
||||
peek_outcomes[i] = Some(outcome);
|
||||
}
|
||||
|
||||
for (i, r) in readers.iter_mut().enumerate() {
|
||||
if errs[i].is_some() {
|
||||
has_err += 1;
|
||||
@@ -702,10 +685,7 @@ async fn list_path_raw_inner(
|
||||
let entry = if let Some(entry) = pending_entries[i].take() {
|
||||
entry
|
||||
} else {
|
||||
let Some(outcome) = peek_outcomes[i].take() else {
|
||||
return Err(DiskError::Unexpected);
|
||||
};
|
||||
match outcome {
|
||||
match peek_with_timeout(&revjob_rx, r, peek_timeout).await {
|
||||
PeekOutcome::Ready(res) => {
|
||||
if let Some(entry) = res {
|
||||
// info!("read entry disk: {}, name: {}", i, entry.name);
|
||||
@@ -1315,36 +1295,6 @@ mod tests {
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
async fn list_path_raw_bounds_multiple_stalled_readers_by_one_peek_deadline() {
|
||||
let peek_timeout = Duration::from_millis(20);
|
||||
let started = tokio::time::Instant::now();
|
||||
let err = list_path_raw(
|
||||
CancellationToken::new(),
|
||||
ListPathRawOptions {
|
||||
disks: vec![None, None, None, None],
|
||||
min_disks: 1,
|
||||
test_reader_behaviors: vec![
|
||||
TestReaderBehavior::Stall,
|
||||
TestReaderBehavior::Stall,
|
||||
TestReaderBehavior::Stall,
|
||||
TestReaderBehavior::Stall,
|
||||
],
|
||||
peek_timeout: Some(peek_timeout),
|
||||
..Default::default()
|
||||
},
|
||||
)
|
||||
.await
|
||||
.expect_err("all stalled readers should fail the listing");
|
||||
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
assert_eq!(
|
||||
started.elapsed(),
|
||||
peek_timeout,
|
||||
"reader deadlines must overlap instead of accumulating once per disk"
|
||||
);
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn list_path_raw_waits_past_producer_stall_for_slow_progressing_reader() {
|
||||
let entry = MetaCacheEntry {
|
||||
|
||||
@@ -229,6 +229,17 @@ pub fn http_resp_to_error_response(
|
||||
err_resp
|
||||
}
|
||||
|
||||
pub fn err_transfer_acceleration_bucket(bucket_name: &str) -> ErrorResponse {
|
||||
ErrorResponse {
|
||||
status_code: StatusCode::BAD_REQUEST,
|
||||
code: S3ErrorCode::InvalidArgument,
|
||||
message: "The name of the bucket used for Transfer Acceleration must be DNS-compliant and must not contain periods ‘.’."
|
||||
.to_string(),
|
||||
bucket_name: bucket_name.to_string(),
|
||||
..Default::default()
|
||||
}
|
||||
}
|
||||
|
||||
pub fn err_entity_too_large(total_size: i64, max_object_size: i64, bucket_name: &str, object_name: &str) -> ErrorResponse {
|
||||
let msg = format!(
|
||||
"Your proposed upload size ‘{}’ exceeds the maximum allowed object size ‘{}’ for single PUT operation.",
|
||||
@@ -284,6 +295,16 @@ pub fn err_invalid_argument(message: &str) -> ErrorResponse {
|
||||
}
|
||||
}
|
||||
|
||||
pub fn err_api_not_supported(message: &str) -> ErrorResponse {
|
||||
ErrorResponse {
|
||||
status_code: StatusCode::NOT_IMPLEMENTED,
|
||||
code: S3ErrorCode::Custom("APINotSupported".into()),
|
||||
message: message.to_string(),
|
||||
request_id: "rustfs".to_string(),
|
||||
..Default::default()
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
mod tests {
|
||||
use super::*;
|
||||
|
||||
@@ -135,10 +135,6 @@ impl Object {
|
||||
Self { ..Default::default() }
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity reader surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn do_get_request(&self, request: &GetRequest) -> Result<GetResponse, std::io::Error> {
|
||||
let _ = request.did_offset_change;
|
||||
let _ = request.offset;
|
||||
@@ -154,20 +150,12 @@ impl Object {
|
||||
))
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn set_offset(&mut self, bytes_read: i64) -> Result<(), std::io::Error> {
|
||||
self.curr_offset += bytes_read;
|
||||
|
||||
Ok(())
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn read(&mut self, b: &[u8]) -> Result<i64, std::io::Error> {
|
||||
let mut read_req = GetRequest {
|
||||
is_read_op: true,
|
||||
@@ -192,10 +180,6 @@ impl Object {
|
||||
Ok(response.size)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn stat(&self) -> Result<ObjectInfo, std::io::Error> {
|
||||
if !self.is_started || !self.object_info_set {
|
||||
let _ = self.do_get_request(&GetRequest {
|
||||
@@ -208,10 +192,6 @@ impl Object {
|
||||
Ok(self.object_info.clone())
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn read_at(&mut self, b: &[u8], offset: i64) -> Result<i64, std::io::Error> {
|
||||
self.curr_offset = offset;
|
||||
|
||||
@@ -239,10 +219,6 @@ impl Object {
|
||||
Ok(response.size)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn seek(&mut self, offset: i64, whence: i64) -> Result<i64, std::io::Error> {
|
||||
if !self.is_started || !self.object_info_set {
|
||||
let seek_req = GetRequest {
|
||||
@@ -277,10 +253,6 @@ impl Object {
|
||||
Ok(self.curr_offset)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity Object reader method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn close(&mut self) -> Result<(), std::io::Error> {
|
||||
self.is_closed = true;
|
||||
Ok(())
|
||||
|
||||
@@ -37,7 +37,7 @@ use crate::client::{
|
||||
api_put_object_common::optimal_part_info,
|
||||
api_put_object_multipart::UploadPartParams,
|
||||
api_s3_datatypes::{CompleteMultipartUpload, CompletePart, ObjectPart},
|
||||
constants::{ISO8601_DATEFORMAT, MAX_MULTIPART_PUT_OBJECT_SIZE, MIN_PART_SIZE},
|
||||
constants::{ISO8601_DATEFORMAT, MAX_MULTIPART_PUT_OBJECT_SIZE, MIN_PART_SIZE, TOTAL_WORKERS},
|
||||
credentials::SignatureType,
|
||||
transition_api::{ReaderImpl, TransitionClient, UploadInfo},
|
||||
utils::{is_amz_header, is_minio_header, is_rustfs_header, is_standard_header, is_storageclass_header},
|
||||
|
||||
@@ -30,6 +30,10 @@ pub fn is_object(reader: &ReaderImpl) -> bool {
|
||||
matches!(reader, ReaderImpl::ObjectBody(_))
|
||||
}
|
||||
|
||||
pub fn is_read_at(reader: ReaderImpl) -> bool {
|
||||
matches!(reader, ReaderImpl::ObjectBody(_))
|
||||
}
|
||||
|
||||
pub fn optimal_part_info(object_size: i64, configured_part_size: u64) -> Result<(i64, i64, i64), std::io::Error> {
|
||||
let unknown_size;
|
||||
let mut object_size = object_size;
|
||||
|
||||
@@ -81,6 +81,18 @@ async fn read_multipart_part(reader: &mut ReaderImpl, want: usize) -> Result<Vec
|
||||
}
|
||||
}
|
||||
|
||||
pub struct UploadedPartRes {
|
||||
pub error: std::io::Error,
|
||||
pub part_num: i64,
|
||||
pub size: i64,
|
||||
pub part: ObjectPart,
|
||||
}
|
||||
|
||||
pub struct UploadPartReq {
|
||||
pub part_num: i64,
|
||||
pub part: ObjectPart,
|
||||
}
|
||||
|
||||
impl TransitionClient {
|
||||
pub async fn put_object_multipart_stream(
|
||||
self: Arc<Self>,
|
||||
|
||||
@@ -29,6 +29,10 @@ use crate::client::utils::base64_decode;
|
||||
|
||||
use super::transition_api;
|
||||
|
||||
pub struct ListAllMyBucketsResult {
|
||||
pub owner: Owner,
|
||||
}
|
||||
|
||||
#[derive(Debug, Default, Serialize, Deserialize)]
|
||||
pub struct CommonPrefix {
|
||||
pub prefix: String,
|
||||
@@ -85,10 +89,6 @@ pub struct ListVersionsResult {
|
||||
pub next_version_id_marker: String,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "fields of a MinIO-parity list result that this port builds but never reads back (backlog#1823)"
|
||||
)]
|
||||
pub struct ListBucketResult {
|
||||
common_prefixes: Vec<CommonPrefix>,
|
||||
contents: Vec<transition_api::ObjectInfo>,
|
||||
@@ -102,10 +102,6 @@ pub struct ListBucketResult {
|
||||
prefix: String,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "fields of a MinIO-parity list result that this port builds but never reads back (backlog#1823)"
|
||||
)]
|
||||
pub struct ListMultipartUploadsResult {
|
||||
bucket: String,
|
||||
key_marker: String,
|
||||
@@ -121,15 +117,16 @@ pub struct ListMultipartUploadsResult {
|
||||
common_prefixes: Vec<CommonPrefix>,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "fields of a MinIO-parity list result that this port builds but never reads back (backlog#1823)"
|
||||
)]
|
||||
pub struct Initiator {
|
||||
id: String,
|
||||
display_name: String,
|
||||
}
|
||||
|
||||
pub struct CopyObjectResult {
|
||||
pub etag: String,
|
||||
pub last_modified: OffsetDateTime,
|
||||
}
|
||||
|
||||
#[derive(Debug, Clone)]
|
||||
pub struct ObjectPart {
|
||||
pub etag: String,
|
||||
@@ -263,7 +260,6 @@ pub struct CompletePart {
|
||||
}
|
||||
|
||||
impl CompletePart {
|
||||
#[allow(dead_code, reason = "MinIO-parity accessor with no caller in this port (backlog#1823)")]
|
||||
fn checksum(&self, t: &ChecksumMode) -> String {
|
||||
match t {
|
||||
ChecksumMode::ChecksumCRC32C => {
|
||||
@@ -288,6 +284,11 @@ impl CompletePart {
|
||||
}
|
||||
}
|
||||
|
||||
pub struct CopyObjectPartResult {
|
||||
pub etag: String,
|
||||
pub last_modified: OffsetDateTime,
|
||||
}
|
||||
|
||||
#[derive(Debug, Default, serde::Serialize)]
|
||||
#[serde(rename = "CompleteMultipartUpload")]
|
||||
pub struct CompleteMultipartUpload {
|
||||
@@ -356,10 +357,10 @@ impl CompleteMultipartUpload {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live via quick_xml::de::from_str in bucket_cache.rs; serde deserialization is not a construction (backlog#1823)"
|
||||
)]
|
||||
pub struct CreateBucketConfiguration {
|
||||
pub location: String,
|
||||
}
|
||||
|
||||
#[derive(serde::Serialize)]
|
||||
pub struct DeleteObject {
|
||||
//api has
|
||||
@@ -367,6 +368,21 @@ pub struct DeleteObject {
|
||||
pub version_id: String,
|
||||
}
|
||||
|
||||
pub struct DeletedObject {
|
||||
//s3s has
|
||||
pub key: String,
|
||||
pub version_id: String,
|
||||
pub deletemarker: bool,
|
||||
pub deletemarker_version_id: String,
|
||||
}
|
||||
|
||||
pub struct NonDeletedObject {
|
||||
pub key: String,
|
||||
pub code: String,
|
||||
pub message: String,
|
||||
pub version_id: String,
|
||||
}
|
||||
|
||||
#[derive(serde::Serialize)]
|
||||
pub struct DeleteMultiObjects {
|
||||
pub quiet: bool,
|
||||
@@ -386,7 +402,6 @@ impl DeleteMultiObjects {
|
||||
Ok(buf)
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "MinIO-parity XML helper with no caller in this port (backlog#1823)")]
|
||||
pub fn unmarshal(buf: &[u8]) -> Result<Self, std::io::Error> {
|
||||
#[derive(Debug, Deserialize)]
|
||||
struct WireDeleteObject {
|
||||
@@ -421,3 +436,8 @@ impl DeleteMultiObjects {
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
pub struct DeleteMultiObjectsResult {
|
||||
pub deleted_objects: Vec<DeletedObject>,
|
||||
pub undeleted_objects: Vec<NonDeletedObject>,
|
||||
}
|
||||
|
||||
@@ -27,24 +27,12 @@ use crate::client::utils::base64_decode;
|
||||
use crate::client::utils::base64_encode;
|
||||
use crate::client::{api_put_object::PutObjectOptions, api_s3_datatypes::ObjectPart};
|
||||
use crate::{disk::DiskAPI, object_api::GetObjectReader};
|
||||
// s3s::header has no CRC64NVME constant yet; the canonical RustFS copy lives
|
||||
// in rustfs-utils' headers module.
|
||||
use rustfs_utils::http::headers::AMZ_CHECKSUM_CRC64NVME;
|
||||
use s3s::header::{
|
||||
X_AMZ_CHECKSUM_ALGORITHM, X_AMZ_CHECKSUM_CRC32, X_AMZ_CHECKSUM_CRC32C, X_AMZ_CHECKSUM_SHA1, X_AMZ_CHECKSUM_SHA256,
|
||||
};
|
||||
|
||||
use enumset::{EnumSet, EnumSetType, enum_set};
|
||||
|
||||
/// One of three deliberately separate checksum registries (backlog#1833):
|
||||
/// this enum is the MinIO-port client's wire vocabulary and stops at the
|
||||
/// standard S3 set (CRC64NVME is its newest member; the RustFS extensions do
|
||||
/// not exist on this client path). The streaming-hash registry lives in
|
||||
/// `rustfs_checksums::ChecksumAlgorithm` (crates/checksums/src/lib.rs) and
|
||||
/// the on-disk xl.meta bitset in `rustfs_rio::ChecksumType`
|
||||
/// (crates/rio/src/checksum.rs, varint bits are append-only). When adding an
|
||||
/// algorithm, extend all three (or record why not) — they do not derive from
|
||||
/// each other.
|
||||
#[derive(Debug, EnumSetType, Default)]
|
||||
#[enumset(repr = "u8")]
|
||||
pub enum ChecksumMode {
|
||||
@@ -69,6 +57,8 @@ lazy_static! {
|
||||
static ref C_ChecksumFullObjectCRC32C: EnumSet<ChecksumMode> =
|
||||
enum_set!(ChecksumMode::ChecksumCRC32C | ChecksumMode::ChecksumFullObject);
|
||||
}
|
||||
const AMZ_CHECKSUM_CRC64NVME: &str = "x-amz-checksum-crc64nvme";
|
||||
|
||||
impl ChecksumMode {
|
||||
//pub const CRC64_NVME_POLYNOMIAL: i64 = 0xad93d23594c93659;
|
||||
|
||||
@@ -365,10 +355,6 @@ mod tests {
|
||||
pub struct Checksum {
|
||||
checksum_type: ChecksumMode,
|
||||
r: Vec<u8>,
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "checksum bookkeeping field kept beside the value it guards (backlog#1823)"
|
||||
)]
|
||||
computed: bool,
|
||||
}
|
||||
|
||||
|
||||
@@ -32,5 +32,8 @@ pub const MAX_MULTIPART_PUT_OBJECT_SIZE: i64 = 1024 * 1024 * 1024 * 1024 * 5;
|
||||
pub const UNSIGNED_PAYLOAD: &str = "UNSIGNED-PAYLOAD";
|
||||
pub const UNSIGNED_PAYLOAD_TRAILER: &str = "STREAMING-UNSIGNED-PAYLOAD-TRAILER";
|
||||
|
||||
pub const TOTAL_WORKERS: i64 = 4;
|
||||
|
||||
pub const SIGN_V4_ALGORITHM: &str = "AWS4-HMAC-SHA256";
|
||||
pub const ISO8601_DATEFORMAT: &[FormatItem<'_>] =
|
||||
format_description!("[year]-[month]-[day]T[hour]:[minute]:[second].[subsecond]Z");
|
||||
|
||||
@@ -67,10 +67,6 @@ impl<P: Provider + Default> Credentials<P> {
|
||||
Ok(self.creds.clone())
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity credential surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn expire(&mut self) {
|
||||
self.force_refresh = true;
|
||||
}
|
||||
@@ -137,10 +133,6 @@ impl Provider for Static {
|
||||
|
||||
#[derive(Debug, Clone, Default)]
|
||||
pub struct STSError {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity STS error detail that this port never reads back (backlog#1823)"
|
||||
)]
|
||||
pub r#type: String,
|
||||
pub code: String,
|
||||
pub message: String,
|
||||
@@ -149,10 +141,6 @@ pub struct STSError {
|
||||
#[derive(Debug, Clone, thiserror::Error)]
|
||||
pub struct ErrorResponse {
|
||||
pub sts_error: STSError,
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity STS error detail that this port never reads back (backlog#1823)"
|
||||
)]
|
||||
pub request_id: String,
|
||||
}
|
||||
|
||||
@@ -170,3 +158,22 @@ impl ErrorResponse {
|
||||
return self.sts_error.message.clone();
|
||||
}
|
||||
}
|
||||
|
||||
pub fn xml_decoder<T>(body: &[u8]) -> Result<T, Error>
|
||||
where
|
||||
for<'de> T: Deserialize<'de>,
|
||||
{
|
||||
match std::str::from_utf8(body) {
|
||||
Ok(xml_body) => quick_xml::de::from_str::<T>(xml_body).map_err(|err| Error::new(ErrorKind::InvalidData, err.to_string())),
|
||||
Err(err) => Err(Error::new(ErrorKind::InvalidData, err.to_string())),
|
||||
}
|
||||
}
|
||||
|
||||
pub fn xml_decode_and_body<T>(body_reader: &[u8]) -> Result<(Vec<u8>, T), std::io::Error>
|
||||
where
|
||||
for<'de> T: Deserialize<'de>,
|
||||
{
|
||||
let body = body_reader.to_vec();
|
||||
let parsed = xml_decoder(&body)?;
|
||||
Ok((body, parsed))
|
||||
}
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: S3 client compatibility models are kept while ECStore callers move to narrower facades.
|
||||
#![allow(dead_code)]
|
||||
|
||||
pub mod admin_handler_utils;
|
||||
pub mod api_error_response;
|
||||
|
||||
@@ -77,6 +77,39 @@ fn part_number_to_rangespec(oi: ObjectInfo, part_number: usize) -> Option<HTTPRa
|
||||
})
|
||||
}
|
||||
|
||||
fn get_compressed_offsets(oi: ObjectInfo, offset: i64) -> (i64, i64, i64, i64, u64) {
|
||||
let mut skip_length: i64 = 0;
|
||||
let mut cumulative_actual_size: i64 = 0;
|
||||
let mut first_part_idx: i64 = 0;
|
||||
let mut compressed_offset: i64 = 0;
|
||||
let mut part_skip: i64 = 0;
|
||||
let mut decrypt_skip: i64 = 0;
|
||||
let mut seq_num: u64 = 0;
|
||||
for (i, part) in oi.parts.iter().enumerate() {
|
||||
cumulative_actual_size += part.actual_size as i64;
|
||||
if cumulative_actual_size <= offset {
|
||||
compressed_offset += part.size as i64;
|
||||
} else {
|
||||
first_part_idx = i as i64;
|
||||
skip_length = cumulative_actual_size - part.actual_size as i64;
|
||||
break;
|
||||
}
|
||||
}
|
||||
skip_length = offset - skip_length;
|
||||
|
||||
let parts: &[ObjectPartInfo] = &oi.parts;
|
||||
if skip_length > 0
|
||||
&& parts.len() > first_part_idx as usize
|
||||
&& parts[first_part_idx as usize].index.as_ref().is_some_and(|idx| idx.len() > 0)
|
||||
{
|
||||
let _ = part_skip;
|
||||
let _ = decrypt_skip;
|
||||
let _ = seq_num;
|
||||
}
|
||||
|
||||
(compressed_offset, part_skip, first_part_idx, decrypt_skip, seq_num)
|
||||
}
|
||||
|
||||
pub fn new_getobjectreader<'a>(
|
||||
rs: &Option<HTTPRangeSpec>,
|
||||
oi: &'a ObjectInfo,
|
||||
|
||||
@@ -23,7 +23,6 @@ const X_OBS_VERSION_ID: &str = "x-obs-version-id";
|
||||
const MAX_REMOTE_VERSION_ID_LEN: usize = 1024;
|
||||
|
||||
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
|
||||
#[allow(dead_code, reason = "bucket versioning states kept as a complete vocabulary (backlog#1823)")]
|
||||
pub(crate) enum BucketVersioningState {
|
||||
Unknown,
|
||||
Disabled,
|
||||
@@ -48,7 +47,6 @@ impl RemoteVersion {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "MinIO-parity accessor with no caller in this port (backlog#1823)")]
|
||||
pub(crate) fn exact_request_id(&self) -> Result<Option<&str>, Error> {
|
||||
match self {
|
||||
Self::Unknown => Err(Error::new(
|
||||
|
||||
@@ -101,10 +101,6 @@ where
|
||||
|
||||
const C_UNKNOWN: i32 = -1;
|
||||
const C_OFFLINE: i32 = 0;
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "reachable only from the unused transition client methods below (backlog#1823)"
|
||||
)]
|
||||
const C_ONLINE: i32 = 1;
|
||||
|
||||
fn invalid_utf8_header_error(scope: &str, header_name: &str) -> std::io::Error {
|
||||
@@ -324,10 +320,6 @@ impl TransitionClient {
|
||||
Ok(client)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client surface with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn endpoint_url(&self) -> Url {
|
||||
self.endpoint_url.clone()
|
||||
}
|
||||
@@ -356,20 +348,12 @@ impl TransitionClient {
|
||||
.to_string())
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn trace_errors_only_off(&self) {
|
||||
if let Ok(mut trace_errors_only) = self.trace_errors_only.lock() {
|
||||
*trace_errors_only = false;
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn trace_off(&self) {
|
||||
if let Ok(mut is_trace_enabled) = self.is_trace_enabled.lock() {
|
||||
*is_trace_enabled = false;
|
||||
@@ -379,20 +363,12 @@ impl TransitionClient {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn set_s3_transfer_accelerate(&self, accelerate_endpoint: &str) {
|
||||
if let Ok(mut endpoint) = self.s3_accelerate_endpoint.lock() {
|
||||
*endpoint = accelerate_endpoint.to_string();
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn set_s3_enable_dual_stack(&self, enabled: bool) {
|
||||
if let Ok(mut dual_stack) = self.s3_dual_stack_enabled.lock() {
|
||||
*dual_stack = enabled;
|
||||
@@ -422,18 +398,10 @@ impl TransitionClient {
|
||||
(hash_algos, hash_sums)
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn is_online(&self) -> bool {
|
||||
!self.is_offline()
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn mark_offline(&self) {
|
||||
self.health_status
|
||||
.compare_exchange(C_ONLINE, C_OFFLINE, Ordering::SeqCst, Ordering::SeqCst);
|
||||
@@ -443,18 +411,10 @@ impl TransitionClient {
|
||||
self.health_status.load(Ordering::SeqCst) == C_OFFLINE
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn health_check(hc_duration: Duration) {
|
||||
let _ = hc_duration;
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "MinIO-parity transition client method with no caller in this port (backlog#1823)"
|
||||
)]
|
||||
fn dump_http(&self, req: &Request<s3s::Body>, resp: &Response<Incoming>) -> Result<(), std::io::Error> {
|
||||
let mut resp_trace: Vec<u8>;
|
||||
|
||||
@@ -1142,7 +1102,6 @@ impl Default for ObjectInfo {
|
||||
}
|
||||
|
||||
impl ObjectInfo {
|
||||
#[allow(dead_code, reason = "MinIO-parity accessor with no caller in this port (backlog#1823)")]
|
||||
pub(crate) fn remote_version(
|
||||
&self,
|
||||
capabilities: ProviderVersionCapabilities,
|
||||
|
||||
@@ -48,6 +48,10 @@ lazy_static! {
|
||||
};
|
||||
}
|
||||
|
||||
pub fn is_standard_query_value(qs_key: &str) -> bool {
|
||||
SUPPORTED_QUERY_VALUES[qs_key]
|
||||
}
|
||||
|
||||
pub fn is_storageclass_header(header_key: &str) -> bool {
|
||||
header_key.to_lowercase() == X_AMZ_STORAGE_CLASS.as_str().to_lowercase()
|
||||
}
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: cluster/RPC migration leaves transport capabilities staged for upcoming owners.
|
||||
#![allow(dead_code)]
|
||||
|
||||
mod control_plane;
|
||||
pub(crate) mod rpc;
|
||||
|
||||
@@ -256,7 +256,6 @@ impl<S> ReplayScopeChannel<S> {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "replay-state probe asserted by this file's tests (backlog#1823)")]
|
||||
fn peer_replay_state(audience: &str) -> PeerReplayState {
|
||||
PEER_REPLAY_STATES
|
||||
.lock()
|
||||
|
||||
@@ -31,7 +31,7 @@ use rustfs_config::{
|
||||
DEFAULT_INTERNODE_DATA_TRANSPORT, ENV_RUSTFS_INTERNODE_DATA_TRANSPORT, INTERNODE_DATA_TRANSPORT_TCP,
|
||||
KNOWN_INTERNODE_DATA_TRANSPORT_BACKENDS,
|
||||
};
|
||||
use rustfs_rio::{ChunkReaderBox, HttpChunkReader, HttpReader, HttpWriter};
|
||||
use rustfs_rio::{HttpReader, HttpWriter};
|
||||
use sha2::{Digest, Sha256};
|
||||
use std::collections::HashMap;
|
||||
use std::future::Future;
|
||||
@@ -43,10 +43,6 @@ use tokio::io::{AsyncReadExt, AsyncWrite};
|
||||
use tokio::sync::OnceCell;
|
||||
use uuid::Uuid;
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live in the cfg(not(test)) half of build_internode_data_transport_from_env (backlog#1823)"
|
||||
)]
|
||||
static INTERNODE_DATA_TRANSPORT: OnceLock<std::result::Result<Arc<dyn InternodeDataTransport>, String>> = OnceLock::new();
|
||||
|
||||
const READ_FILE_STREAM_PATH: &str = "/rustfs/rpc/read_file_stream";
|
||||
@@ -138,10 +134,6 @@ fn put_file_capability_status_is_legacy(status: u16) -> bool {
|
||||
}
|
||||
|
||||
#[derive(Debug, Clone, Copy, Eq, PartialEq)]
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "capability-negotiation seam; constructed only by transport test doubles (backlog#1823)"
|
||||
)]
|
||||
pub struct InternodeDataTransportCapabilities {
|
||||
/// Backend can open a streaming remote disk reader.
|
||||
pub streaming_read: bool,
|
||||
@@ -158,10 +150,6 @@ pub struct InternodeDataTransportCapabilities {
|
||||
}
|
||||
|
||||
impl InternodeDataTransportCapabilities {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "capability-negotiation seam; used by transport test doubles (backlog#1823)"
|
||||
)]
|
||||
pub const fn tcp_http() -> Self {
|
||||
Self {
|
||||
streaming_read: true,
|
||||
@@ -233,11 +221,6 @@ pub struct NsScannerCapabilityRequest {
|
||||
#[async_trait]
|
||||
pub trait InternodeDataTransport: Send + Sync + std::fmt::Debug {
|
||||
async fn open_read(&self, request: ReadStreamRequest) -> Result<FileReader>;
|
||||
/// Opens an owned-chunk stream when this transport can retain receive-buffer
|
||||
/// ownership. `None` preserves the established `open_read` fallback.
|
||||
async fn open_read_chunks(&self, _request: ReadStreamRequest) -> Result<Option<ChunkReaderBox>> {
|
||||
Ok(None)
|
||||
}
|
||||
async fn open_write(&self, request: WriteStreamRequest) -> Result<FileWriter>;
|
||||
async fn open_walk_dir(&self, request: WalkDirStreamRequest) -> Result<FileReader>;
|
||||
async fn open_ns_scanner(&self, _request: NsScannerStreamRequest) -> Result<FileReader> {
|
||||
@@ -246,12 +229,7 @@ pub trait InternodeDataTransport: Send + Sync + std::fmt::Debug {
|
||||
async fn probe_ns_scanner(&self, _request: NsScannerCapabilityRequest) -> Result<Uuid> {
|
||||
Err(Error::MethodNotAllowed)
|
||||
}
|
||||
// Interface facet nobody calls yet: every transport implements both, but no
|
||||
// caller negotiates on them. Kept for the internode transport split
|
||||
// (backlog#1350); deleting them would delete the seam and six impls.
|
||||
#[allow(dead_code, reason = "unused capability-negotiation facet (backlog#1823)")]
|
||||
fn name(&self) -> &'static str;
|
||||
#[allow(dead_code, reason = "unused capability-negotiation facet (backlog#1823)")]
|
||||
fn capabilities(&self) -> InternodeDataTransportCapabilities;
|
||||
}
|
||||
|
||||
@@ -269,15 +247,6 @@ impl InternodeDataTransport for TcpHttpInternodeDataTransport {
|
||||
))
|
||||
}
|
||||
|
||||
async fn open_read_chunks(&self, request: ReadStreamRequest) -> Result<Option<ChunkReaderBox>> {
|
||||
let url = build_read_file_stream_url(&request);
|
||||
let mut headers = json_headers();
|
||||
build_auth_headers(&url, &Method::GET, &mut headers)?;
|
||||
Ok(Some(Box::new(
|
||||
HttpChunkReader::new_with_stall_timeout(url, Method::GET, headers, None, request.stall_timeout).await?,
|
||||
)))
|
||||
}
|
||||
|
||||
async fn open_write(&self, request: WriteStreamRequest) -> Result<FileWriter> {
|
||||
let server_epoch = self.put_file_auth_capability(&request.endpoint).await?;
|
||||
let nonce = server_epoch.map(|_| Uuid::new_v4());
|
||||
@@ -687,10 +656,6 @@ fn build_internode_data_transport_result(
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "live in the cfg(test) half of build_internode_data_transport_from_env, which bypasses the process static (backlog#1823)"
|
||||
)]
|
||||
pub fn build_internode_data_transport(configured_transport: Option<&str>) -> Result<Arc<dyn InternodeDataTransport>> {
|
||||
build_internode_data_transport_result(configured_transport).map_err(Error::other)
|
||||
}
|
||||
|
||||
@@ -248,16 +248,6 @@ fn decode_remote_version_state_capability(expected_member: &str, result: &[u8])
|
||||
Ok(server_epoch)
|
||||
}
|
||||
|
||||
fn decode_cross_pool_fence_capability(expected_member: &str, result: &[u8]) -> Result<(u32, Uuid)> {
|
||||
let version = result
|
||||
.get(..4)
|
||||
.and_then(|value| value.try_into().ok())
|
||||
.map(u32::from_be_bytes)
|
||||
.ok_or_else(|| Error::other("peer returned an invalid cross-pool fence capability version"))?;
|
||||
let epoch = decode_remote_version_state_capability(expected_member, &result[4..])?;
|
||||
Ok((version, epoch))
|
||||
}
|
||||
|
||||
#[derive(Clone, Debug)]
|
||||
pub struct PeerLiveEventsBatch {
|
||||
pub events: Vec<u8>,
|
||||
@@ -1298,16 +1288,6 @@ impl PeerRestClient {
|
||||
Ok((self.topology_member.clone(), epoch))
|
||||
}
|
||||
|
||||
pub async fn probe_cross_pool_fence(&self, topology_fingerprint: String) -> Result<(String, u32, Uuid)> {
|
||||
let mut probe = rustfs_protos::CROSS_POOL_FENCE_CAPABILITY_PROBE_PREFIX.to_vec();
|
||||
probe.extend_from_slice(Uuid::new_v4().as_bytes());
|
||||
let result = self
|
||||
.heal_control(rustfs_protos::HEAL_CONTROL_PROTOCOL_VERSION, topology_fingerprint, probe)
|
||||
.await?;
|
||||
let (supported_version, epoch) = decode_cross_pool_fence_capability(&self.topology_member, &result)?;
|
||||
Ok((self.topology_member.clone(), supported_version, epoch))
|
||||
}
|
||||
|
||||
pub async fn load_bucket_metadata(&self, bucket: &str, scanner_maintenance_change: bool) -> Result<()> {
|
||||
self.finalize_result(
|
||||
async {
|
||||
@@ -2758,24 +2738,6 @@ mod tests {
|
||||
assert!(decode_remote_version_state_capability("node-a:9000", &nil).is_err());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn cross_pool_fence_capability_decoder_fails_closed() {
|
||||
let epoch = Uuid::new_v4();
|
||||
let result = rustfs_protos::encode_cross_pool_fence_capability(1, "node-a:9000", epoch.as_bytes())
|
||||
.expect("small capability response should encode");
|
||||
assert_eq!(
|
||||
decode_cross_pool_fence_capability("node-a:9000", &result).expect("valid capability should decode"),
|
||||
(1, epoch)
|
||||
);
|
||||
for malformed in [&[][..], &[0, 0, 0][..], &result[..result.len() - 1]] {
|
||||
assert!(decode_cross_pool_fence_capability("node-a:9000", malformed).is_err());
|
||||
}
|
||||
assert!(decode_cross_pool_fence_capability("node-b:9000", &result).is_err());
|
||||
let nil = rustfs_protos::encode_cross_pool_fence_capability(1, "node-a:9000", Uuid::nil().as_bytes())
|
||||
.expect("small capability response should encode");
|
||||
assert!(decode_cross_pool_fence_capability("node-a:9000", &nil).is_err());
|
||||
}
|
||||
|
||||
struct TierMutationResponseFixture<'a> {
|
||||
version: u32,
|
||||
phase: TierMutationRpcPhase,
|
||||
|
||||
@@ -854,6 +854,7 @@ impl PeerS3Client for LocalPeerS3Client {
|
||||
|
||||
#[derive(Debug)]
|
||||
pub struct RemotePeerS3Client {
|
||||
pub node: Option<Node>,
|
||||
pub pools: Option<Vec<usize>>,
|
||||
addr: String,
|
||||
/// Health tracker for connection monitoring
|
||||
@@ -885,6 +886,7 @@ impl RemotePeerS3Client {
|
||||
pub fn new(node: Option<Node>, pools: Option<Vec<usize>>) -> Self {
|
||||
let addr = node.as_ref().map(|v| v.url.to_string()).unwrap_or_default();
|
||||
let client = Self {
|
||||
node,
|
||||
pools,
|
||||
addr,
|
||||
health: Arc::new(DiskHealthTracker::new()),
|
||||
@@ -903,6 +905,10 @@ impl RemotePeerS3Client {
|
||||
.map_err(|err| Error::other(format!("can not get client, err: {err}")))
|
||||
}
|
||||
|
||||
pub fn get_addr(&self) -> String {
|
||||
self.addr.clone()
|
||||
}
|
||||
|
||||
/// Start health monitoring for the remote peer
|
||||
fn start_health_monitoring(&self) {
|
||||
let health = Arc::clone(&self.health);
|
||||
@@ -1202,10 +1208,6 @@ impl PeerS3Client for RemotePeerS3Client {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "local bucket-heal path reached only by this file's tests (backlog#1823)"
|
||||
)]
|
||||
pub async fn heal_bucket_local(bucket: &str, opts: &HealOpts) -> Result<HealResultItem> {
|
||||
let disks = clone_drives().await;
|
||||
heal_bucket_local_on_disks(bucket, opts, disks).await
|
||||
@@ -1402,10 +1404,6 @@ pub(crate) async fn heal_bucket_local_on_disks(
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "reached only through heal_bucket_local, which only tests call (backlog#1823)"
|
||||
)]
|
||||
async fn clone_drives() -> Vec<Option<DiskStore>> {
|
||||
runtime_sources::local_disk_entries().await
|
||||
}
|
||||
@@ -1587,7 +1585,15 @@ mod tests {
|
||||
}
|
||||
|
||||
fn test_remote_peer(addr: &str) -> RemotePeerS3Client {
|
||||
let node = Node {
|
||||
url: url::Url::parse(addr).expect("test peer URL should parse"),
|
||||
pools: vec![0],
|
||||
is_local: false,
|
||||
grid_host: addr.to_string(),
|
||||
};
|
||||
|
||||
RemotePeerS3Client {
|
||||
node: Some(node),
|
||||
pools: Some(vec![0]),
|
||||
addr: addr.to_string(),
|
||||
health: Arc::new(DiskHealthTracker::new()),
|
||||
|
||||
@@ -522,33 +522,6 @@ impl RemoteDisk {
|
||||
}
|
||||
}
|
||||
|
||||
async fn open_read_chunks_with_retry(&self, request: ReadStreamRequest) -> Result<Option<rustfs_rio::ChunkReaderBox>> {
|
||||
let mut attempt = 1;
|
||||
let mut last_retry_classification = None;
|
||||
loop {
|
||||
match self.data_transport.open_read_chunks(request.clone()).await {
|
||||
Ok(reader) => {
|
||||
if attempt > 1
|
||||
&& let Some(classification) = last_retry_classification
|
||||
{
|
||||
crate::cluster::rpc::runtime_sources::record_remote_disk_open_read_retry_success(classification);
|
||||
}
|
||||
return Ok(reader);
|
||||
}
|
||||
Err(err) if attempt < REMOTE_DISK_OPEN_READ_MAX_ATTEMPTS && Self::is_retryable_open_read_error(&err) => {
|
||||
if let Some(classification) = err.internode_http_error_kind() {
|
||||
let classification = classification.metric_label();
|
||||
crate::cluster::rpc::runtime_sources::record_remote_disk_open_read_retry(classification);
|
||||
last_retry_classification = Some(classification);
|
||||
}
|
||||
tokio::time::sleep(REMOTE_DISK_OPEN_READ_RETRY_BACKOFF).await;
|
||||
attempt += 1;
|
||||
}
|
||||
Err(err) => return Err(err),
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
pub fn record_capacity_probe(&self, total: u64, used: u64, free: u64) {
|
||||
self.health.record_capacity_probe(total, used, free);
|
||||
}
|
||||
@@ -873,49 +846,31 @@ impl RemoteDisk {
|
||||
/// default to 1 (see [`internode_idempotent_read_retries`]). MUST NOT be used for write/lock
|
||||
/// RPCs — those must never auto-retry (quorum/idempotency safety). The `operation` closure is
|
||||
/// re-invoked per attempt, so it must be `Fn` (rebuild the request from borrowed inputs, do not
|
||||
/// move captured state out). Attempts and backoff share one total timeout budget.
|
||||
/// move captured state out).
|
||||
async fn execute_read_with_retry<T, F, Fut>(&self, op: &'static str, operation: F, timeout_duration: Duration) -> Result<T>
|
||||
where
|
||||
F: Fn() -> Fut,
|
||||
Fut: std::future::Future<Output = Result<T>>,
|
||||
{
|
||||
let deadline = (!timeout_duration.is_zero()).then(|| {
|
||||
time::Instant::now()
|
||||
.checked_add(timeout_duration)
|
||||
.unwrap_or_else(|| time::sleep(timeout_duration).deadline())
|
||||
});
|
||||
let max_retries = internode_idempotent_read_retries();
|
||||
let mut attempt = 0usize;
|
||||
loop {
|
||||
let attempt_timeout = deadline
|
||||
.map(|deadline| deadline.saturating_duration_since(time::Instant::now()))
|
||||
.unwrap_or(Duration::ZERO);
|
||||
if deadline.is_some() && attempt_timeout.is_zero() {
|
||||
self.record_timeout(op, timeout_duration);
|
||||
return Err(DiskError::Timeout);
|
||||
}
|
||||
|
||||
// Only the final attempt marks the disk faulty / evicts the channel. Earlier retries
|
||||
// ignore the failure, so a transient error cannot flip the disk into a faulty
|
||||
// short-circuit (which would defeat the retry) or over-count failures.
|
||||
let health_action = if attempt >= max_retries {
|
||||
FailureHealthAction::MarkFailure
|
||||
} else {
|
||||
FailureHealthAction::IgnoreFailure
|
||||
};
|
||||
match self
|
||||
.execute_with_timeout_for_op_and_health_action(op, &operation, attempt_timeout, health_action)
|
||||
.execute_with_timeout_for_op_and_health_action(op, &operation, timeout_duration, health_action)
|
||||
.await
|
||||
{
|
||||
Err(err) if attempt < max_retries && is_network_like_disk_error(&err) => {
|
||||
if matches!(err, DiskError::Timeout) && deadline.is_some_and(|deadline| time::Instant::now() >= deadline) {
|
||||
self.mark_faulty("read_operation_deadline");
|
||||
return Err(err);
|
||||
}
|
||||
attempt += 1;
|
||||
let backoff = REMOTE_DISK_READ_RETRY_BASE_BACKOFF
|
||||
.saturating_mul(1u32 << u32::try_from(attempt - 1).unwrap_or(4).min(4));
|
||||
if deadline.is_some_and(|deadline| deadline.saturating_duration_since(time::Instant::now()) <= backoff) {
|
||||
attempt = max_retries;
|
||||
continue;
|
||||
}
|
||||
debug!(
|
||||
endpoint = %self.endpoint,
|
||||
addr = %self.addr,
|
||||
@@ -923,17 +878,7 @@ impl RemoteDisk {
|
||||
attempt,
|
||||
"retrying idempotent read-only RPC after transient network error"
|
||||
);
|
||||
if let Some(deadline) = deadline {
|
||||
if time::timeout_at(deadline, time::sleep(backoff)).await.is_err() {
|
||||
self.record_timeout(op, timeout_duration);
|
||||
return Err(DiskError::Timeout);
|
||||
}
|
||||
} else {
|
||||
time::sleep(backoff).await;
|
||||
}
|
||||
if self.health.is_faulty() {
|
||||
return Err(DiskError::FaultyDisk);
|
||||
}
|
||||
tokio::time::sleep(backoff).await;
|
||||
}
|
||||
other => return other,
|
||||
}
|
||||
@@ -1012,35 +957,32 @@ impl RemoteDisk {
|
||||
operation_result
|
||||
}
|
||||
Err(_) => {
|
||||
self.record_timeout(op, timeout_duration);
|
||||
// Timeout occurred, mark disk as potentially faulty
|
||||
counter!(
|
||||
"rustfs_drive_op_timeout_total",
|
||||
"endpoint" => self.endpoint.to_string(),
|
||||
"op" => op.to_string()
|
||||
)
|
||||
.increment(1);
|
||||
if failure_health_action == FailureHealthAction::MarkFailure {
|
||||
self.mark_faulty_and_evict("operation_timeout").await;
|
||||
}
|
||||
warn!(
|
||||
event = EVENT_REMOTE_DISK_RPC,
|
||||
component = LOG_COMPONENT_ECSTORE,
|
||||
subsystem = LOG_SUBSYSTEM_REMOTE_DISK,
|
||||
endpoint = %self.endpoint,
|
||||
addr = %self.addr,
|
||||
op,
|
||||
timeout_ms = timeout_duration.as_millis(),
|
||||
state = "timeout",
|
||||
"Remote disk operation timed out"
|
||||
);
|
||||
Err(DiskError::Timeout)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
fn record_timeout(&self, op: &'static str, timeout_duration: Duration) {
|
||||
counter!(
|
||||
"rustfs_drive_op_timeout_total",
|
||||
"endpoint" => self.endpoint.to_string(),
|
||||
"op" => op.to_string()
|
||||
)
|
||||
.increment(1);
|
||||
warn!(
|
||||
event = EVENT_REMOTE_DISK_RPC,
|
||||
component = LOG_COMPONENT_ECSTORE,
|
||||
subsystem = LOG_SUBSYSTEM_REMOTE_DISK,
|
||||
endpoint = %self.endpoint,
|
||||
addr = %self.addr,
|
||||
op,
|
||||
timeout_ms = timeout_duration.as_millis(),
|
||||
state = "timeout",
|
||||
"Remote disk operation timed out"
|
||||
);
|
||||
}
|
||||
|
||||
async fn handle_network_like_error<T>(
|
||||
&self,
|
||||
op: &'static str,
|
||||
@@ -1074,7 +1016,7 @@ impl RemoteDisk {
|
||||
}
|
||||
}
|
||||
|
||||
fn mark_faulty(&self, reason: &'static str) -> bool {
|
||||
async fn mark_faulty_and_evict(&self, reason: &'static str) {
|
||||
let previous_state = self.runtime_state();
|
||||
let transitioned_to_offline = self.mark_suspect_or_offline(reason);
|
||||
let state = self.runtime_state();
|
||||
@@ -1111,12 +1053,6 @@ impl RemoteDisk {
|
||||
"Remote disk marked suspect"
|
||||
);
|
||||
}
|
||||
}
|
||||
state != previous_state
|
||||
}
|
||||
|
||||
async fn mark_faulty_and_evict(&self, reason: &'static str) {
|
||||
if self.mark_faulty(reason) {
|
||||
counter!(
|
||||
"rustfs_drive_connection_evict_total",
|
||||
"endpoint" => self.endpoint.to_string(),
|
||||
@@ -1359,71 +1295,6 @@ fn validate_decoded_file_info(file_info: &FileInfo) -> Result<()> {
|
||||
file_info.validate_for_metadata_read().map_err(Into::into)
|
||||
}
|
||||
|
||||
impl RemoteDisk {
|
||||
#[tracing::instrument(level = "trace", skip_all)]
|
||||
pub(crate) async fn rename_data_borrowed(
|
||||
&self,
|
||||
src_volume: &str,
|
||||
src_path: &str,
|
||||
fi: &FileInfo,
|
||||
dst_volume: &str,
|
||||
dst_path: &str,
|
||||
) -> Result<RenameDataResp> {
|
||||
trace!(
|
||||
event = EVENT_REMOTE_DISK_RPC,
|
||||
component = LOG_COMPONENT_ECSTORE,
|
||||
subsystem = LOG_SUBSYSTEM_REMOTE_DISK,
|
||||
endpoint = %self.endpoint,
|
||||
src_volume,
|
||||
src_path,
|
||||
dst_volume,
|
||||
dst_path,
|
||||
op = "rename_data",
|
||||
state = "started",
|
||||
"Remote disk RPC started"
|
||||
);
|
||||
|
||||
self.execute_with_timeout_for_op(
|
||||
"rename_data",
|
||||
|| async {
|
||||
let file_info = compat_json(fi)?;
|
||||
let file_info_bin = encode_file_info_msgpack(fi)?;
|
||||
let mut client = self
|
||||
.get_client()
|
||||
.await
|
||||
.map_err(|err| Error::other(format!("can not get client, err: {err}")))?;
|
||||
let mut request = Request::new(RenameDataRequest {
|
||||
disk: self.endpoint.to_string(),
|
||||
src_volume: src_volume.to_string(),
|
||||
src_path: src_path.to_string(),
|
||||
file_info,
|
||||
dst_volume: dst_volume.to_string(),
|
||||
dst_path: dst_path.to_string(),
|
||||
file_info_bin: file_info_bin.into(),
|
||||
});
|
||||
let canonical_body = rustfs_protos::canonical_rename_data_request_body(request.get_ref());
|
||||
attach_mutation_body_digest(&mut request, canonical_body, "rename_data")?;
|
||||
|
||||
let response = client.rename_data(request).await?.into_inner();
|
||||
|
||||
if !response.success {
|
||||
return Err(response.error.unwrap_or_default().into());
|
||||
}
|
||||
|
||||
let rename_data_resp = decode_msgpack_or_json::<RenameDataResp>(
|
||||
&response.rename_data_resp_bin,
|
||||
&response.rename_data_resp,
|
||||
"RenameDataResp",
|
||||
)?;
|
||||
|
||||
Ok(rename_data_resp)
|
||||
},
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
}
|
||||
|
||||
#[async_trait::async_trait]
|
||||
impl DiskAPI for RemoteDisk {
|
||||
#[tracing::instrument(level = "trace", skip_all)]
|
||||
@@ -2197,7 +2068,7 @@ impl DiskAPI for RemoteDisk {
|
||||
|
||||
Ok(file_info)
|
||||
},
|
||||
get_drive_metadata_timeout(),
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
@@ -2349,8 +2220,58 @@ impl DiskAPI for RemoteDisk {
|
||||
dst_volume: &str,
|
||||
dst_path: &str,
|
||||
) -> Result<RenameDataResp> {
|
||||
self.rename_data_borrowed(src_volume, src_path, &fi, dst_volume, dst_path)
|
||||
.await
|
||||
trace!(
|
||||
event = EVENT_REMOTE_DISK_RPC,
|
||||
component = LOG_COMPONENT_ECSTORE,
|
||||
subsystem = LOG_SUBSYSTEM_REMOTE_DISK,
|
||||
endpoint = %self.endpoint,
|
||||
src_volume,
|
||||
src_path,
|
||||
dst_volume,
|
||||
dst_path,
|
||||
op = "rename_data",
|
||||
state = "started",
|
||||
"Remote disk RPC started"
|
||||
);
|
||||
|
||||
self.execute_with_timeout_for_op(
|
||||
"rename_data",
|
||||
|| async {
|
||||
let file_info = compat_json(&fi)?;
|
||||
let file_info_bin = encode_file_info_msgpack(&fi)?;
|
||||
let mut client = self
|
||||
.get_client()
|
||||
.await
|
||||
.map_err(|err| Error::other(format!("can not get client, err: {err}")))?;
|
||||
let mut request = Request::new(RenameDataRequest {
|
||||
disk: self.endpoint.to_string(),
|
||||
src_volume: src_volume.to_string(),
|
||||
src_path: src_path.to_string(),
|
||||
file_info,
|
||||
dst_volume: dst_volume.to_string(),
|
||||
dst_path: dst_path.to_string(),
|
||||
file_info_bin: file_info_bin.into(),
|
||||
});
|
||||
let canonical_body = rustfs_protos::canonical_rename_data_request_body(request.get_ref());
|
||||
attach_mutation_body_digest(&mut request, canonical_body, "rename_data")?;
|
||||
|
||||
let response = client.rename_data(request).await?.into_inner();
|
||||
|
||||
if !response.success {
|
||||
return Err(response.error.unwrap_or_default().into());
|
||||
}
|
||||
|
||||
let rename_data_resp = decode_msgpack_or_json::<RenameDataResp>(
|
||||
&response.rename_data_resp_bin,
|
||||
&response.rename_data_resp,
|
||||
"RenameDataResp",
|
||||
)?;
|
||||
|
||||
Ok(rename_data_resp)
|
||||
},
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
|
||||
#[tracing::instrument(level = "trace", skip_all)]
|
||||
@@ -2496,30 +2417,6 @@ impl DiskAPI for RemoteDisk {
|
||||
.await
|
||||
}
|
||||
|
||||
async fn read_file_stream_chunks(
|
||||
&self,
|
||||
volume: &str,
|
||||
path: &str,
|
||||
offset: usize,
|
||||
length: usize,
|
||||
) -> Result<Option<rustfs_rio::ChunkReaderBox>> {
|
||||
if self.health.is_faulty() {
|
||||
return Err(DiskError::FaultyDisk);
|
||||
}
|
||||
let disk = self.disk_ref().await;
|
||||
let stall_timeout = get_object_disk_read_timeout();
|
||||
self.open_read_chunks_with_retry(ReadStreamRequest {
|
||||
endpoint: self.endpoint.grid_host(),
|
||||
disk,
|
||||
volume: volume.to_string(),
|
||||
path: path.to_string(),
|
||||
offset,
|
||||
length,
|
||||
stall_timeout: (!stall_timeout.is_zero()).then_some(stall_timeout),
|
||||
})
|
||||
.await
|
||||
}
|
||||
|
||||
/// Buffered read for remote disks.
|
||||
/// The transport stream is collected into owned Bytes for caller sharing.
|
||||
#[tracing::instrument(level = "trace", skip_all)]
|
||||
@@ -5197,452 +5094,6 @@ mod tests {
|
||||
);
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_reset_during_backoff_preserves_recovery() {
|
||||
let remote_disk = Arc::new(new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await);
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let first_attempt = Arc::new(tokio::sync::Notify::new());
|
||||
let started = time::Instant::now();
|
||||
|
||||
let task_disk = Arc::clone(&remote_disk);
|
||||
let task_attempts = Arc::clone(&attempts);
|
||||
let task_first_attempt = Arc::clone(&first_attempt);
|
||||
let task = tokio::spawn(async move {
|
||||
task_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
move || {
|
||||
let attempt = task_attempts.fetch_add(1, Ordering::SeqCst);
|
||||
let first_attempt = Arc::clone(&task_first_attempt);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
time::sleep(Duration::from_millis(20)).await;
|
||||
first_attempt.notify_one();
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionRefused,
|
||||
"connection refused",
|
||||
)));
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
});
|
||||
|
||||
first_attempt.notified().await;
|
||||
tokio::task::yield_now().await;
|
||||
remote_disk.health.reset_for_store_init_retry(&remote_disk.endpoint);
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
task.await
|
||||
.expect("retry task should finish")
|
||||
.expect("the retry should succeed after the health reset");
|
||||
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(started.elapsed(), Duration::from_millis(70));
|
||||
assert_eq!(
|
||||
remote_disk.health.waiting_count(),
|
||||
0,
|
||||
"health reset must not underflow the waiting counter"
|
||||
);
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Online);
|
||||
assert!(
|
||||
runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await,
|
||||
"a recovered channel must survive the retry backoff"
|
||||
);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_still_retries_within_shared_deadline() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
|
||||
remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
let attempt = attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionReset,
|
||||
"connection reset",
|
||||
)));
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
.expect("a retry that fits the shared deadline should succeed");
|
||||
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Online);
|
||||
assert!(runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_uses_remaining_budget_for_final_attempt() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
let attempt = attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
time::sleep(Duration::from_millis(20)).await;
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionRefused,
|
||||
"connection refused",
|
||||
)));
|
||||
}
|
||||
std::future::pending::<Result<()>>().await
|
||||
}
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
.expect_err("the final retry should consume only the remaining total budget");
|
||||
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(started.elapsed(), Duration::from_millis(100));
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
assert!(!runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_uses_final_attempt_at_exact_backoff_boundary() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
let attempt = attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
time::sleep(Duration::from_millis(50)).await;
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionRefused,
|
||||
"connection refused",
|
||||
)));
|
||||
}
|
||||
std::future::pending::<Result<()>>().await
|
||||
}
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
.expect_err("the exact backoff boundary should be reserved for a final attempt");
|
||||
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(started.elapsed(), Duration::from_millis(100));
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_uses_final_attempt_below_backoff_budget() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
let attempt = attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
time::sleep(Duration::from_millis(80)).await;
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionRefused,
|
||||
"connection refused",
|
||||
)));
|
||||
}
|
||||
std::future::pending::<Result<()>>().await
|
||||
}
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
.expect_err("remaining budget below backoff should be reserved for a final attempt");
|
||||
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(started.elapsed(), Duration::from_millis(100));
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_zero_timeout_disables_the_deadline() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
|
||||
remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
let attempt = attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionReset,
|
||||
"connection reset",
|
||||
)));
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
},
|
||||
Duration::ZERO,
|
||||
)
|
||||
.await
|
||||
.expect("zero timeout should allow a retry without a deadline");
|
||||
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 2);
|
||||
assert_eq!(started.elapsed(), REMOTE_DISK_READ_RETRY_BASE_BACKOFF);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_accepts_max_metadata_timeout() {
|
||||
temp_env::async_with_vars([(rustfs_config::ENV_DRIVE_METADATA_TIMEOUT_SECS, Some(u64::MAX.to_string()))], async {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
|
||||
remote_disk
|
||||
.execute_read_with_retry("read_version", || async { Ok::<(), Error>(()) }, get_drive_metadata_timeout())
|
||||
.await
|
||||
.expect("the maximum configured metadata timeout must not panic");
|
||||
|
||||
remote_disk.cancel_token.cancel();
|
||||
})
|
||||
.await;
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_zero_retries_runs_once() {
|
||||
temp_env::async_with_vars([(rustfs_config::ENV_INTERNODE_IDEMPOTENT_READ_RETRIES, Some("0"))], async {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async {
|
||||
Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionReset,
|
||||
"connection reset",
|
||||
)))
|
||||
}
|
||||
},
|
||||
Duration::from_secs(1),
|
||||
)
|
||||
.await
|
||||
.expect_err("zero retries should return the first network error");
|
||||
|
||||
assert!(matches!(err, DiskError::Io(ref io_err) if io_err.kind() == std_io::ErrorKind::ConnectionReset));
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 1);
|
||||
assert_eq!(started.elapsed(), Duration::ZERO);
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
assert!(!runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await);
|
||||
remote_disk.cancel_token.cancel();
|
||||
})
|
||||
.await;
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_attempt_timeout_marks_health_without_evicting() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let recorder = crate::test_metrics::CapturingRecorder::default();
|
||||
let _recorder_guard = metrics::set_default_local_recorder(&recorder);
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
attempts.fetch_add(1, Ordering::SeqCst);
|
||||
std::future::pending::<Result<()>>()
|
||||
},
|
||||
Duration::from_millis(100),
|
||||
)
|
||||
.await
|
||||
.expect_err("an in-flight attempt that consumes the deadline should time out");
|
||||
|
||||
assert_eq!(err, DiskError::Timeout);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 1);
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
assert!(runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await);
|
||||
assert_eq!(
|
||||
recorder.counter_value(
|
||||
"rustfs_drive_op_timeout_total",
|
||||
&[
|
||||
("endpoint", remote_disk.endpoint.to_string().as_str()),
|
||||
("op", "read_version")
|
||||
]
|
||||
),
|
||||
1
|
||||
);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_does_not_retry_business_errors() {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async { Err::<(), Error>(DiskError::FileNotFound) }
|
||||
},
|
||||
Duration::from_secs(1),
|
||||
)
|
||||
.await
|
||||
.expect_err("business errors should be returned directly");
|
||||
|
||||
assert_eq!(err, DiskError::FileNotFound);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 1);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_honors_configured_retry_count() {
|
||||
temp_env::async_with_vars([(rustfs_config::ENV_INTERNODE_IDEMPOTENT_READ_RETRIES, Some("2"))], async {
|
||||
let remote_disk = new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await;
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let started = time::Instant::now();
|
||||
let channel = TonicEndpoint::from_shared(remote_disk.addr.clone())
|
||||
.expect("remote disk address should parse")
|
||||
.connect_lazy();
|
||||
runtime_sources::cache_test_node_channel(remote_disk.addr.clone(), channel).await;
|
||||
|
||||
let err = remote_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
|| {
|
||||
attempts.fetch_add(1, Ordering::SeqCst);
|
||||
async {
|
||||
Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionReset,
|
||||
"connection reset",
|
||||
)))
|
||||
}
|
||||
},
|
||||
Duration::from_secs(1),
|
||||
)
|
||||
.await
|
||||
.expect_err("exhausted retries should return the last network error");
|
||||
|
||||
assert!(matches!(err, DiskError::Io(ref io_err) if io_err.kind() == std_io::ErrorKind::ConnectionReset));
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 3);
|
||||
assert_eq!(started.elapsed(), Duration::from_millis(150));
|
||||
assert_eq!(remote_disk.runtime_state(), RuntimeDriveHealthState::Suspect);
|
||||
assert!(!runtime_sources::test_node_channel_is_cached(&remote_disk.addr).await);
|
||||
remote_disk.cancel_token.cancel();
|
||||
})
|
||||
.await;
|
||||
}
|
||||
|
||||
#[tokio::test(start_paused = true)]
|
||||
#[serial(remote_disk_read_retry)]
|
||||
async fn execute_read_with_retry_stops_when_disk_turns_offline_during_backoff() {
|
||||
let remote_disk = Arc::new(new_remote_disk_with_transport(Arc::new(RecordingInternodeDataTransport::default())).await);
|
||||
let attempts = Arc::new(std::sync::atomic::AtomicUsize::new(0));
|
||||
let first_attempt = Arc::new(tokio::sync::Notify::new());
|
||||
let task_disk = Arc::clone(&remote_disk);
|
||||
let task_attempts = Arc::clone(&attempts);
|
||||
let task_first_attempt = Arc::clone(&first_attempt);
|
||||
|
||||
let task = tokio::spawn(async move {
|
||||
task_disk
|
||||
.execute_read_with_retry(
|
||||
"read_version",
|
||||
move || {
|
||||
let attempt = task_attempts.fetch_add(1, Ordering::SeqCst);
|
||||
let first_attempt = Arc::clone(&task_first_attempt);
|
||||
async move {
|
||||
if attempt == 0 {
|
||||
first_attempt.notify_one();
|
||||
return Err::<(), Error>(DiskError::Io(std_io::Error::new(
|
||||
std_io::ErrorKind::ConnectionReset,
|
||||
"connection reset",
|
||||
)));
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
},
|
||||
Duration::from_secs(1),
|
||||
)
|
||||
.await
|
||||
});
|
||||
|
||||
first_attempt.notified().await;
|
||||
tokio::task::yield_now().await;
|
||||
remote_disk
|
||||
.health
|
||||
.force_runtime_state_for_test(RuntimeDriveHealthState::Offline);
|
||||
time::advance(REMOTE_DISK_READ_RETRY_BASE_BACKOFF).await;
|
||||
let err = task
|
||||
.await
|
||||
.expect("retry task should finish")
|
||||
.expect_err("an offline disk must stop before the next attempt");
|
||||
|
||||
assert_eq!(err, DiskError::FaultyDisk);
|
||||
assert_eq!(attempts.load(Ordering::SeqCst), 1);
|
||||
remote_disk.cancel_token.cancel();
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn test_execute_with_timeout_evicts_cached_connection() {
|
||||
let addr = "http://127.0.0.1:59991".to_string();
|
||||
@@ -6152,40 +5603,6 @@ mod tests {
|
||||
accept_task.abort();
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn read_version_uses_the_metadata_timeout_on_a_stalled_peer() {
|
||||
runtime_sources::ensure_test_rpc_secret();
|
||||
let Some((base_addr, accept_task)) = spawn_stalled_grpc_peer().await else {
|
||||
return;
|
||||
};
|
||||
let remote_disk = remote_disk_for_addr(&base_addr).await;
|
||||
|
||||
temp_env::async_with_vars(
|
||||
[
|
||||
(rustfs_config::ENV_DRIVE_METADATA_TIMEOUT_SECS, Some("1")),
|
||||
(rustfs_config::ENV_DRIVE_MAX_TIMEOUT_DURATION, Some("10")),
|
||||
],
|
||||
async {
|
||||
let started = time::Instant::now();
|
||||
let err = tokio::time::timeout(
|
||||
Duration::from_secs(5),
|
||||
remote_disk.read_version("bucket", "bucket", "object", "", &ReadOptions::default()),
|
||||
)
|
||||
.await
|
||||
.expect("read_version must use the shorter metadata deadline")
|
||||
.expect_err("a stalled peer must fail read_version");
|
||||
|
||||
assert!(matches!(err, DiskError::Timeout), "expected the metadata deadline to fire, got {err:?}");
|
||||
assert!(started.elapsed() >= Duration::from_millis(900));
|
||||
assert!(started.elapsed() < Duration::from_secs(2));
|
||||
},
|
||||
)
|
||||
.await;
|
||||
|
||||
remote_disk.cancel_token.cancel();
|
||||
accept_task.abort();
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn delete_volume_bounds_the_wait_on_a_stalled_peer() {
|
||||
runtime_sources::ensure_test_rpc_secret();
|
||||
|
||||
@@ -48,6 +48,10 @@ impl RemoteClient {
|
||||
Self { addr: endpoint }
|
||||
}
|
||||
|
||||
pub fn from_url(url: url::Url) -> Self {
|
||||
Self { addr: url.to_string() }
|
||||
}
|
||||
|
||||
fn build_ping_request() -> PingRequest {
|
||||
let mut fbb = flatbuffers::FlatBufferBuilder::new();
|
||||
let payload = fbb.create_vector(b"health-check");
|
||||
|
||||
@@ -46,6 +46,7 @@ use rustfs_config::{
|
||||
SCANNER_SUB_SYS,
|
||||
};
|
||||
use rustfs_filemeta::FileInfo;
|
||||
use rustfs_utils::path::SLASH_SEPARATOR;
|
||||
use serde_json::{Map, Value};
|
||||
use std::collections::{HashMap, HashSet};
|
||||
use std::sync::LazyLock;
|
||||
@@ -199,6 +200,8 @@ pub const STORAGE_CLASS_SUB_SYS: &str = "storage_class";
|
||||
|
||||
pub const COMMA_SEPARATED_LISTS: &[&str] = &[rustfs_config::oidc::OIDC_SCOPES, rustfs_config::oidc::OIDC_OTHER_AUDIENCES];
|
||||
|
||||
static CONFIG_BUCKET: LazyLock<String> = LazyLock::new(|| format!("{RUSTFS_META_BUCKET}{SLASH_SEPARATOR}{CONFIG_PREFIX}"));
|
||||
|
||||
type ServerConfigDecryptFn = crate::bucket::migration::LegacyBlobDecryptFn;
|
||||
|
||||
static SERVER_CONFIG_DECRYPT_FN: LazyLock<RwLock<Option<ServerConfigDecryptFn>>> = LazyLock::new(|| RwLock::new(None));
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: configuration migration keeps legacy subsystem definitions available behind this module.
|
||||
#![allow(dead_code)]
|
||||
|
||||
mod audit;
|
||||
pub mod com;
|
||||
|
||||
@@ -101,7 +101,6 @@ const DEFAULT_RRS_STORAGE_CLASS: &str = "EC:1";
|
||||
const ZERO_SET_DRIVE_COUNT_ERROR: &str = "set drive count must be greater than zero";
|
||||
|
||||
pub static DEFAULT_INLINE_BLOCK: usize = 128 * 1024;
|
||||
const DEFAULT_INLINE_OBJECT_BUDGET: usize = 2 * DEFAULT_INLINE_BLOCK;
|
||||
|
||||
pub static DEFAULT_KVS: LazyLock<KVS> = LazyLock::new(|| {
|
||||
let kvs = vec![
|
||||
@@ -151,8 +150,6 @@ pub struct Config {
|
||||
optimize: Option<String>,
|
||||
inline_block: usize,
|
||||
initialized: bool,
|
||||
#[serde(default, skip_serializing_if = "std::ops::Not::not")]
|
||||
inline_block_explicit: bool,
|
||||
#[serde(skip)]
|
||||
standard_parities: Vec<PoolParity>,
|
||||
#[serde(skip)]
|
||||
@@ -189,10 +186,6 @@ impl Config {
|
||||
/// A topology-bound lookup fails closed for unknown drive counts and for
|
||||
/// deserialized legacy configurations that have no pool topology. Legacy
|
||||
/// callers retain scalar compatibility through [`Self::get_parity_for_sc`].
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "per-set parity resolution asserted by this file's tests (backlog#1823)"
|
||||
)]
|
||||
pub(crate) fn parity_for_sc(&self, sc: &str, drives_per_set: usize) -> Option<usize> {
|
||||
if !self.initialized {
|
||||
return None;
|
||||
@@ -240,19 +233,17 @@ impl Config {
|
||||
.map(|(pool_index, pool)| (pool_index, pool.drives_per_set))
|
||||
}
|
||||
|
||||
pub fn should_inline(&self, shard_size: i64, data_shards: usize, versioned: bool) -> bool {
|
||||
if shard_size < 0 || data_shards == 0 {
|
||||
pub fn should_inline(&self, shard_size: i64, versioned: bool) -> bool {
|
||||
if shard_size < 0 {
|
||||
return false;
|
||||
}
|
||||
|
||||
let shard_size = shard_size as usize;
|
||||
// Keep the historical two-data-shard object budget while preventing
|
||||
// wider EC layouts from multiplying the maximum inline object size.
|
||||
let inline_block = if self.initialized && self.inline_block_explicit {
|
||||
self.inline_block
|
||||
} else {
|
||||
(DEFAULT_INLINE_OBJECT_BUDGET / data_shards).min(DEFAULT_INLINE_BLOCK)
|
||||
};
|
||||
|
||||
let mut inline_block = DEFAULT_INLINE_BLOCK;
|
||||
if self.initialized {
|
||||
inline_block = self.inline_block;
|
||||
}
|
||||
|
||||
if versioned {
|
||||
shard_size <= inline_block / 8
|
||||
@@ -401,7 +392,6 @@ fn lookup_config_for_pools_with_env(
|
||||
}
|
||||
|
||||
let optimize = overrides.optimize;
|
||||
let inline_block_explicit = overrides.inline_block.is_some();
|
||||
let inline_block = if let Some(value) = overrides.inline_block {
|
||||
let block = value
|
||||
.parse::<bytesize::ByteSize>()
|
||||
@@ -434,7 +424,6 @@ fn lookup_config_for_pools_with_env(
|
||||
optimize,
|
||||
inline_block,
|
||||
initialized: true,
|
||||
inline_block_explicit,
|
||||
standard_parities,
|
||||
rrs_parities,
|
||||
})
|
||||
@@ -552,26 +541,22 @@ mod tests {
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn should_inline_scales_default_threshold_by_data_shards() {
|
||||
let config = lookup_config_for_pools_with_env(&KVS::new(), &[3, 12], no_env_overrides())
|
||||
.expect("default inline policy should resolve for EC2+1 and EC8+4");
|
||||
fn should_inline_preserves_exact_default_shard_boundaries() {
|
||||
let config = Config::default();
|
||||
|
||||
for (case, shard_size, data_shards, versioned, expected) in [
|
||||
("EC2+1 unversioned exact", 128 * 1024, 2, false, true),
|
||||
("EC2+1 unversioned above", 128 * 1024 + 1, 2, false, false),
|
||||
("EC2+1 versioned exact", 16 * 1024, 2, true, true),
|
||||
("EC2+1 versioned above", 16 * 1024 + 1, 2, true, false),
|
||||
("EC8+4 unversioned exact", 32 * 1024, 8, false, true),
|
||||
("EC8+4 unversioned above", 32 * 1024 + 1, 8, false, false),
|
||||
("EC8+4 versioned exact", 4 * 1024, 8, true, true),
|
||||
("EC8+4 versioned above", 4 * 1024 + 1, 8, true, false),
|
||||
("negative", -1, 2, false, false),
|
||||
("zero data shards", 0, 0, false, false),
|
||||
for (case, shard_size, versioned, expected) in [
|
||||
("unversioned below", 128 * 1024 - 1, false, true),
|
||||
("unversioned exact", 128 * 1024, false, true),
|
||||
("unversioned above", 128 * 1024 + 1, false, false),
|
||||
("versioned below", 16 * 1024 - 1, true, true),
|
||||
("versioned exact", 16 * 1024, true, true),
|
||||
("versioned above", 16 * 1024 + 1, true, false),
|
||||
("negative", -1, false, false),
|
||||
] {
|
||||
assert_eq!(
|
||||
config.should_inline(shard_size, data_shards, versioned),
|
||||
config.should_inline(shard_size, versioned),
|
||||
expected,
|
||||
"{case}: shard_size={shard_size}, data_shards={data_shards}, versioned={versioned}"
|
||||
"{case}: shard_size={shard_size}, versioned={versioned}"
|
||||
);
|
||||
}
|
||||
}
|
||||
@@ -592,28 +577,13 @@ mod tests {
|
||||
let shard_size = erasure.shard_file_size(object_size);
|
||||
assert_eq!(shard_size, expected_shard_size, "{case}: object_size={object_size}");
|
||||
assert_eq!(
|
||||
config.should_inline(shard_size, erasure.data_shards, versioned),
|
||||
config.should_inline(shard_size, versioned),
|
||||
expected,
|
||||
"{case}: object_size={object_size}, shard_size={shard_size}, versioned={versioned}"
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn explicit_inline_block_preserves_fixed_per_shard_rollback() {
|
||||
let overrides = StorageClassEnvOverrides {
|
||||
inline_block: Some("128KiB".to_string()),
|
||||
..Default::default()
|
||||
};
|
||||
let config = lookup_config_for_pools_with_env(&KVS::new(), &[12], overrides)
|
||||
.expect("explicit inline block should resolve for EC8+4");
|
||||
|
||||
assert!(config.should_inline(128 * 1024, 8, false));
|
||||
assert!(!config.should_inline(128 * 1024 + 1, 8, false));
|
||||
assert!(config.should_inline(16 * 1024, 8, true));
|
||||
assert!(!config.should_inline(16 * 1024 + 1, 8, true));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn write_capability_contract_only_accepts_implemented_layouts() {
|
||||
assert_eq!(SUPPORTED_WRITE_CLASSES, [STANDARD, RRS]);
|
||||
@@ -807,7 +777,6 @@ mod tests {
|
||||
let encoded = serde_json::to_string(&cfg).expect("config should serialize");
|
||||
assert!(!encoded.contains("standard_parities"));
|
||||
assert!(!encoded.contains("rrs_parities"));
|
||||
assert!(!encoded.contains("inline_block_explicit"));
|
||||
|
||||
let decoded: Config = serde_json::from_str(&encoded).expect("legacy scalar config should deserialize");
|
||||
assert_eq!(decoded.get_parity_for_sc(STANDARD), Some(2));
|
||||
@@ -817,25 +786,6 @@ mod tests {
|
||||
assert!(validate_parity(0, 0).is_err());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn explicit_inline_block_survives_config_round_trip() {
|
||||
let cfg = lookup_config_for_pools_with_env(
|
||||
&KVS::new(),
|
||||
&[12],
|
||||
StorageClassEnvOverrides {
|
||||
inline_block: Some("128KiB".to_string()),
|
||||
..Default::default()
|
||||
},
|
||||
)
|
||||
.expect("explicit inline block should resolve");
|
||||
assert!(cfg.should_inline(100 * 1024, 8, false));
|
||||
|
||||
let encoded = serde_json::to_string(&cfg).expect("config should serialize");
|
||||
assert!(encoded.contains("\"inline_block_explicit\":true"));
|
||||
let decoded: Config = serde_json::from_str(&encoded).expect("explicit inline config should deserialize");
|
||||
assert!(decoded.should_inline(100 * 1024, 8, false));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn lookup_config_reads_rrs_from_class_rrs_key() {
|
||||
// Regression: kvs.get(RRS) used RRS="REDUCED_REDUNDANCY" instead of
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: pool coordination helpers are being migrated behind runtime owners.
|
||||
#![allow(dead_code)]
|
||||
|
||||
pub(crate) mod pools;
|
||||
pub(crate) mod sets;
|
||||
|
||||
@@ -226,7 +226,6 @@ fn ensure_decommission_start_rebalance_meta_allowed(meta: Option<&RebalanceMeta>
|
||||
ensure_decommission_not_rebalancing(meta.is_some_and(is_rebalance_conflicting_with_decommission))
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "leader precondition asserted by this file's tests (backlog#1823)")]
|
||||
fn ensure_local_decommission_pool_leaders(endpoints: &EndpointServerPools, indices: &[usize]) -> Result<()> {
|
||||
for idx in indices {
|
||||
ensure_local_decommission_pool_leader(endpoints, *idx)?;
|
||||
@@ -1059,19 +1058,11 @@ fn should_cleanup_decommission_source_entry(decommissioned: usize, total_version
|
||||
}
|
||||
|
||||
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "terminal-state classification asserted by this file's tests (backlog#1823)"
|
||||
)]
|
||||
enum DecommissionTerminalState {
|
||||
Completed,
|
||||
Failed,
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "terminal-state classification asserted by this file's tests (backlog#1823)"
|
||||
)]
|
||||
fn classify_decommission_terminal_state(failed_items_present: bool) -> DecommissionTerminalState {
|
||||
if failed_items_present {
|
||||
DecommissionTerminalState::Failed
|
||||
@@ -2275,19 +2266,15 @@ fn decommission_delete_marker_opts(
|
||||
version: &rustfs_filemeta::FileInfo,
|
||||
version_id: Option<String>,
|
||||
src_pool_idx: usize,
|
||||
expected_bucket_incarnation_id: Option<uuid::Uuid>,
|
||||
) -> ObjectOptions {
|
||||
let version_suspended = version.version_id.is_none() && version_id.is_none();
|
||||
ObjectOptions {
|
||||
versioned: !version_suspended,
|
||||
version_suspended,
|
||||
version_id: version_id.or_else(|| version_suspended.then(|| uuid::Uuid::nil().to_string())),
|
||||
versioned: true,
|
||||
version_id,
|
||||
mod_time: version.mod_time,
|
||||
src_pool_idx,
|
||||
data_movement: true,
|
||||
delete_marker: true,
|
||||
skip_decommissioned: true,
|
||||
expected_bucket_incarnation_id,
|
||||
delete_replication: version
|
||||
.replication_state_internal
|
||||
.as_ref()
|
||||
@@ -2312,7 +2299,6 @@ fn decommission_remote_tiered_opts(
|
||||
version: &rustfs_filemeta::FileInfo,
|
||||
version_id: Option<String>,
|
||||
src_pool_idx: usize,
|
||||
expected_bucket_incarnation_id: Option<uuid::Uuid>,
|
||||
) -> ObjectOptions {
|
||||
ObjectOptions {
|
||||
versioned: version_id.is_some(),
|
||||
@@ -2321,9 +2307,6 @@ fn decommission_remote_tiered_opts(
|
||||
user_defined: version.metadata.clone(),
|
||||
src_pool_idx,
|
||||
data_movement: true,
|
||||
include_part_checksums: true,
|
||||
http_preconditions: Some(crate::data_movement::data_movement_target_precondition()),
|
||||
expected_bucket_incarnation_id,
|
||||
..Default::default()
|
||||
}
|
||||
}
|
||||
@@ -2822,7 +2805,6 @@ impl ECStore {
|
||||
lifecycle_config: Option<BucketLifecycleConfiguration>,
|
||||
object_lock_config: Option<ObjectLockConfiguration>,
|
||||
replication_config: Option<(ReplicationConfiguration, OffsetDateTime)>,
|
||||
expected_bucket_incarnation_id: Option<uuid::Uuid>,
|
||||
) -> Result<()> {
|
||||
debug!(
|
||||
event = EVENT_DECOMMISSION_ENTRY,
|
||||
@@ -2852,11 +2834,6 @@ impl ECStore {
|
||||
}
|
||||
decommission_cancel_signal_result(rx.is_cancelled())?;
|
||||
|
||||
let bucket_incarnation_fence = match expected_bucket_incarnation_id {
|
||||
Some(expected) => Some(self.acquire_bucket_incarnation_fence(&bucket, expected).await?),
|
||||
None => None,
|
||||
};
|
||||
|
||||
let mut fivs = load_decommission_entry_exact_versions(&set, &entry, &bucket, "file_info_versions").await?;
|
||||
|
||||
fivs.versions
|
||||
@@ -2917,7 +2894,7 @@ impl ECStore {
|
||||
.delete_object(
|
||||
bucket.as_str(),
|
||||
&version.name,
|
||||
decommission_delete_marker_opts(version, version_id.clone(), idx, expected_bucket_incarnation_id),
|
||||
decommission_delete_marker_opts(version, version_id.clone(), idx),
|
||||
)
|
||||
.await
|
||||
{
|
||||
@@ -3007,7 +2984,7 @@ impl ECStore {
|
||||
bucket.as_str(),
|
||||
&version.name,
|
||||
version,
|
||||
&decommission_remote_tiered_opts(version, version_id.clone(), idx, expected_bucket_incarnation_id),
|
||||
&decommission_remote_tiered_opts(version, version_id.clone(), idx),
|
||||
)
|
||||
.await
|
||||
{
|
||||
@@ -3079,11 +3056,7 @@ impl ECStore {
|
||||
)
|
||||
.await?;
|
||||
|
||||
if let Err(err) = self
|
||||
.clone()
|
||||
.decommission_object(idx, bucket, rd, expected_bucket_incarnation_id)
|
||||
.await
|
||||
{
|
||||
if let Err(err) = self.clone().decommission_object(idx, bucket, rd).await {
|
||||
if is_decommission_copy_cleanup_safe_error(&err) {
|
||||
ignore = true;
|
||||
cleanup_ignored = true;
|
||||
@@ -3160,9 +3133,6 @@ impl ECStore {
|
||||
}
|
||||
|
||||
if should_cleanup_decommission_source_entry(decommissioned, fivs.versions.len(), expired) {
|
||||
if bucket_incarnation_fence.as_ref().is_some_and(|guard| guard.is_lock_lost()) {
|
||||
return Err(Error::other("decommission bucket incarnation fence was lost before source cleanup"));
|
||||
}
|
||||
decommission_cancel_signal_result(rx.is_cancelled())?;
|
||||
|
||||
self.save_decommission_entry_progress_stage(
|
||||
@@ -3187,12 +3157,6 @@ impl ECStore {
|
||||
entry.name.as_str(),
|
||||
&fivs,
|
||||
&cleanup_preflight_allowed_missing,
|
||||
data_movement::SourceCleanupBucketFence {
|
||||
expected_incarnation_id: expected_bucket_incarnation_id,
|
||||
lifecycle_guard: bucket_incarnation_fence
|
||||
.as_ref()
|
||||
.and_then(|guard| guard.namespace_lock_guard()),
|
||||
},
|
||||
"decommission",
|
||||
)
|
||||
.await
|
||||
@@ -3304,11 +3268,6 @@ impl ECStore {
|
||||
let mut lifecycle_config = None;
|
||||
let mut object_lock_config = None;
|
||||
let mut replication_config = None;
|
||||
let expected_bucket_incarnation_id = if bi.name == RUSTFS_META_BUCKET {
|
||||
None
|
||||
} else {
|
||||
Some(self.bucket_incarnation_id_from_disk(&bi.name).await?)
|
||||
};
|
||||
|
||||
if bi.name != RUSTFS_META_BUCKET {
|
||||
let _ = resolve_decommission_optional_bucket_config_result(
|
||||
@@ -3362,7 +3321,6 @@ impl ECStore {
|
||||
let lifecycle_config = lifecycle_config.clone();
|
||||
let object_lock_config = object_lock_config.clone();
|
||||
let replication_config = replication_config.clone();
|
||||
let expected_bucket_incarnation_id = expected_bucket_incarnation_id;
|
||||
let entry_error = entry_error.clone();
|
||||
let callback_rx = callback_rx.clone();
|
||||
|
||||
@@ -3425,7 +3383,6 @@ impl ECStore {
|
||||
lifecycle_config,
|
||||
object_lock_config,
|
||||
replication_config,
|
||||
expected_bucket_incarnation_id,
|
||||
)
|
||||
.await
|
||||
{
|
||||
@@ -4211,24 +4168,10 @@ impl ECStore {
|
||||
}
|
||||
|
||||
#[tracing::instrument(skip(self, rd))]
|
||||
async fn decommission_object(
|
||||
self: Arc<Self>,
|
||||
pool_idx: usize,
|
||||
bucket: String,
|
||||
rd: GetObjectReader,
|
||||
expected_bucket_incarnation_id: Option<uuid::Uuid>,
|
||||
) -> Result<()> {
|
||||
async fn decommission_object(self: Arc<Self>, pool_idx: usize, bucket: String, rd: GetObjectReader) -> Result<()> {
|
||||
warn!("decommission_object: start {} {}", &bucket, &rd.object_info.name);
|
||||
let object_name = rd.object_info.name.clone();
|
||||
let result = data_movement::migrate_object(
|
||||
self,
|
||||
pool_idx,
|
||||
bucket.clone(),
|
||||
rd,
|
||||
expected_bucket_incarnation_id,
|
||||
"decommission_object",
|
||||
)
|
||||
.await;
|
||||
let result = data_movement::migrate_object(self, pool_idx, bucket.clone(), rd, "decommission_object").await;
|
||||
if result.is_ok() {
|
||||
warn!("decommission_object: migrated {} {}", &bucket, &object_name);
|
||||
}
|
||||
@@ -4404,8 +4347,7 @@ mod tests {
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
let incarnation = uuid::Uuid::new_v4();
|
||||
let opts = decommission_delete_marker_opts(&version, Some("version-id".to_string()), 7, Some(incarnation));
|
||||
let opts = decommission_delete_marker_opts(&version, Some("version-id".to_string()), 7);
|
||||
let replication = opts.delete_replication.expect("replication state should be preserved");
|
||||
|
||||
assert!(opts.versioned);
|
||||
@@ -4415,25 +4357,11 @@ mod tests {
|
||||
assert_eq!(opts.src_pool_idx, 7);
|
||||
assert_eq!(opts.version_id.as_deref(), Some("version-id"));
|
||||
assert_eq!(opts.mod_time, Some(mod_time));
|
||||
assert_eq!(opts.expected_bucket_incarnation_id, Some(incarnation));
|
||||
assert_eq!(replication.replica_status, ReplicationStatusType::Replica);
|
||||
assert!(replication.delete_marker);
|
||||
assert_eq!(replication.replicate_decision_str, "existing");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn decommission_delete_marker_opts_preserves_suspended_null_version() {
|
||||
let version = rustfs_filemeta::FileInfo {
|
||||
deleted: true,
|
||||
..Default::default()
|
||||
};
|
||||
let opts = decommission_delete_marker_opts(&version, None, 7, None);
|
||||
|
||||
assert!(!opts.versioned);
|
||||
assert!(opts.version_suspended);
|
||||
assert_eq!(opts.version_id.as_deref(), Some(uuid::Uuid::nil().to_string().as_str()));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_decommission_object_migration_read_opts_are_raw_data_movement() {
|
||||
let opts = decommission_object_migration_read_opts(Some("vid-1".to_string()));
|
||||
@@ -4455,8 +4383,7 @@ mod tests {
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
let incarnation = uuid::Uuid::new_v4();
|
||||
let opts = decommission_remote_tiered_opts(&version, Some("version-id".to_string()), 9, Some(incarnation));
|
||||
let opts = decommission_remote_tiered_opts(&version, Some("version-id".to_string()), 9);
|
||||
|
||||
assert!(opts.versioned);
|
||||
assert!(opts.data_movement);
|
||||
@@ -4464,9 +4391,6 @@ mod tests {
|
||||
assert_eq!(opts.version_id.as_deref(), Some("version-id"));
|
||||
assert_eq!(opts.mod_time, Some(mod_time));
|
||||
assert_eq!(opts.user_defined.get("x-amz-meta-key").map(String::as_str), Some("value"));
|
||||
assert!(opts.include_part_checksums);
|
||||
assert!(opts.http_preconditions.is_some());
|
||||
assert_eq!(opts.expected_bucket_incarnation_id, Some(incarnation));
|
||||
}
|
||||
|
||||
#[test]
|
||||
|
||||
+253
-1564
File diff suppressed because it is too large
Load Diff
@@ -12,16 +12,6 @@
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
//! Per-disk usage snapshots persisted under the metadata bucket.
|
||||
//!
|
||||
//! **Nothing calls into this module.** It landed complete with tests in #5307
|
||||
//! (2026-07-27) and its aggregation entry point,
|
||||
//! [`crate::data_usage::aggregate_local_snapshots`], has never had a caller in
|
||||
//! the tree's history. The live data-usage path is
|
||||
//! `load_data_usage_from_backend` / `store_data_usage_in_backend`. The items
|
||||
//! below therefore carry individual `dead_code` allows rather than a module
|
||||
//! blanket, so the gap stays greppable until it is either wired up or removed.
|
||||
|
||||
use crate::data_usage::BucketUsageInfo;
|
||||
use crate::disk::RUSTFS_META_BUCKET;
|
||||
use crate::error::{Error, Result};
|
||||
@@ -36,12 +26,10 @@ pub const DATA_USAGE_DIR: &str = "datausage";
|
||||
/// Directory used to store incremental scan state files under the metadata bucket.
|
||||
pub const DATA_USAGE_STATE_DIR: &str = "datausage/state";
|
||||
/// Snapshot file format version, allows forward compatibility if the structure evolves.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub const LOCAL_USAGE_SNAPSHOT_VERSION: u32 = 1;
|
||||
|
||||
/// Additional metadata describing which disk produced the snapshot.
|
||||
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub struct LocalUsageSnapshotMeta {
|
||||
/// Disk UUID stored as a string for simpler serialization.
|
||||
pub disk_id: String,
|
||||
@@ -55,7 +43,6 @@ pub struct LocalUsageSnapshotMeta {
|
||||
|
||||
/// Usage snapshot produced by a single disk.
|
||||
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub struct LocalUsageSnapshot {
|
||||
/// Format version recorded in the snapshot.
|
||||
pub format_version: u32,
|
||||
@@ -77,7 +64,6 @@ pub struct LocalUsageSnapshot {
|
||||
pub objects_total_size: u64,
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
impl LocalUsageSnapshot {
|
||||
/// Create an empty snapshot with the default format version filled in.
|
||||
pub fn new(meta: LocalUsageSnapshotMeta) -> Self {
|
||||
@@ -113,13 +99,11 @@ impl LocalUsageSnapshot {
|
||||
}
|
||||
|
||||
/// Build the snapshot file name `<disk-id>.json`.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub fn snapshot_file_name(disk_id: &str) -> String {
|
||||
format!("{disk_id}.json")
|
||||
}
|
||||
|
||||
/// Build the object path relative to `RUSTFS_META_BUCKET`, e.g. `datausage/<disk-id>.json`.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub fn snapshot_object_path(disk_id: &str) -> String {
|
||||
format!("{}/{}", DATA_USAGE_DIR, snapshot_file_name(disk_id))
|
||||
}
|
||||
@@ -135,13 +119,11 @@ pub fn data_usage_state_dir(root: &Path) -> PathBuf {
|
||||
}
|
||||
|
||||
/// Build the absolute path to the snapshot file for the provided disk ID.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub fn snapshot_path(root: &Path, disk_id: &str) -> PathBuf {
|
||||
data_usage_dir(root).join(snapshot_file_name(disk_id))
|
||||
}
|
||||
|
||||
/// Read a snapshot from disk if it exists.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub async fn read_snapshot(root: &Path, disk_id: &str) -> Result<Option<LocalUsageSnapshot>> {
|
||||
let path = snapshot_path(root, disk_id);
|
||||
match fs::read(&path).await {
|
||||
@@ -156,7 +138,6 @@ pub async fn read_snapshot(root: &Path, disk_id: &str) -> Result<Option<LocalUsa
|
||||
}
|
||||
|
||||
/// Persist a snapshot to disk, creating directories as needed and overwriting any existing file.
|
||||
#[allow(dead_code, reason = "unwired local usage-snapshot feature; see module docs (backlog#1823)")]
|
||||
pub async fn write_snapshot(root: &Path, disk_id: &str, snapshot: &LocalUsageSnapshot) -> Result<()> {
|
||||
let dir = data_usage_dir(root);
|
||||
fs::create_dir_all(&dir).await.map_err(Error::other)?;
|
||||
|
||||
@@ -13,6 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: scanner/data-usage state is partially migrated and still owns staged cache helpers.
|
||||
#![allow(dead_code)]
|
||||
|
||||
pub mod local_snapshot;
|
||||
|
||||
@@ -33,8 +34,8 @@ use crate::{
|
||||
pub use local_snapshot::{LocalUsageSnapshot, read_snapshot as read_local_snapshot, snapshot_path};
|
||||
use rustfs_data_usage::{
|
||||
BucketTargetUsageInfo, BucketUsageInfo, CompressionTotalInfo, DATA_USAGE_OBJECT_NAME, DATA_USAGE_OBSERVED_OBJECT_NAME,
|
||||
DataUsageCache, DataUsageInfo, DiskUsageStatus, LEGACY_DATA_USAGE_OBJECT_NAME, SizeHistogram, VersionsHistogram,
|
||||
observed_data_usage_is_newer,
|
||||
DataUsageCache, DataUsageEntry, DataUsageInfo, DiskUsageStatus, LEGACY_DATA_USAGE_OBJECT_NAME, SizeHistogram, SizeSummary,
|
||||
VersionsHistogram, observed_data_usage_is_newer,
|
||||
};
|
||||
use rustfs_io_metrics::record_system_path_failure;
|
||||
use rustfs_utils::path::SLASH_SEPARATOR;
|
||||
@@ -54,6 +55,7 @@ use tracing::{debug, error, info, instrument};
|
||||
// Data usage storage constants
|
||||
pub const DATA_USAGE_ROOT: &str = SLASH_SEPARATOR;
|
||||
const DATA_COMPRESSION_TOTAL_NAME: &str = ".compression.json";
|
||||
const DATA_USAGE_BLOOM_NAME: &str = ".bloomcycle.bin";
|
||||
pub const DATA_USAGE_CACHE_NAME: &str = ".usage-cache.bin";
|
||||
const DATA_USAGE_CACHE_TTL_SECS: u64 = 30;
|
||||
const LIVE_BUCKET_USAGE_MAX_ENTRIES: u64 = 1024;
|
||||
@@ -311,6 +313,11 @@ lazy_static::lazy_static! {
|
||||
LEGACY_DATA_USAGE_OBJECT_NAME
|
||||
);
|
||||
static ref LEGACY_DATA_USAGE_OBJ_BACKUP_PATH: String = format!("{}.bkp", LEGACY_DATA_USAGE_OBJ_NAME_PATH.as_str());
|
||||
pub static ref DATA_USAGE_BLOOM_NAME_PATH: String = format!("{}{}{}",
|
||||
crate::disk::BUCKET_META_PREFIX,
|
||||
SLASH_SEPARATOR,
|
||||
DATA_USAGE_BLOOM_NAME
|
||||
);
|
||||
pub static ref DATA_COMPRESSION_TOTAL_NAME_PATH: String = format!("{}{}{}",
|
||||
crate::disk::BUCKET_META_PREFIX,
|
||||
SLASH_SEPARATOR,
|
||||
@@ -851,10 +858,6 @@ async fn resolve_loaded_snapshot_pair_with_source(
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "primary/backup snapshot fallback asserted by this file's tests (backlog#1823)"
|
||||
)]
|
||||
async fn resolve_loaded_snapshot(
|
||||
primary: Result<Vec<u8>, Error>,
|
||||
backup: impl Future<Output = Result<Vec<u8>, Error>>,
|
||||
@@ -1184,10 +1187,6 @@ pub async fn invalidate_admin_data_usage_snapshot_cache() {
|
||||
}
|
||||
|
||||
/// Aggregate usage information from local disk snapshots.
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "reached only through aggregate_local_snapshots, which has no caller (backlog#1823)"
|
||||
)]
|
||||
fn merge_snapshot(aggregated: &mut DataUsageInfo, mut snapshot: LocalUsageSnapshot, latest_update: &mut Option<SystemTime>) {
|
||||
if let Some(update) = snapshot.last_update
|
||||
&& latest_update.is_none_or(|current| update > current)
|
||||
@@ -1221,10 +1220,6 @@ fn merge_snapshot(aggregated: &mut DataUsageInfo, mut snapshot: LocalUsageSnapsh
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "entry point of the local usage-snapshot feature, which has had no caller since it landed in #5307 (backlog#1823)"
|
||||
)]
|
||||
pub async fn aggregate_local_snapshots(store: Arc<ECStore>) -> Result<(Vec<DiskUsageStatus>, DataUsageInfo), Error> {
|
||||
let mut aggregated = DataUsageInfo::default();
|
||||
let mut latest_update: Option<SystemTime> = None;
|
||||
@@ -1360,7 +1355,7 @@ impl BucketUsageAccumulator {
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
let object_size = quota_object_size(object)?;
|
||||
let object_size = object.size.max(0) as u64;
|
||||
self.current_live_versions = self.current_live_versions.saturating_add(1);
|
||||
self.size_histogram.add(object_size);
|
||||
self.total_size = self.total_size.saturating_add(object_size);
|
||||
@@ -1390,31 +1385,6 @@ impl BucketUsageAccumulator {
|
||||
}
|
||||
}
|
||||
|
||||
pub fn quota_object_size(object: &ObjectInfo) -> Result<u64, Error> {
|
||||
let logical_size = u64::try_from(object.get_actual_size().map_err(Error::other)?).map_err(|_| Error::PartMissingOrCorrupt)?;
|
||||
let persisted_part_size = if object.parts.is_empty() {
|
||||
u64::try_from(object.size).map_err(|_| Error::PartMissingOrCorrupt)?
|
||||
} else {
|
||||
object.parts.iter().try_fold(0_u64, |total, part| {
|
||||
// Compressed streaming objects persist -1 when the transformed
|
||||
// part size is unknown. The physical part size remains a valid
|
||||
// quota floor; reject only non-negative values that overflow.
|
||||
let actual_size = if part.actual_size < 0 {
|
||||
if object.is_compressed() {
|
||||
0
|
||||
} else {
|
||||
return Err(Error::PartMissingOrCorrupt);
|
||||
}
|
||||
} else {
|
||||
u64::try_from(part.actual_size).map_err(|_| Error::PartMissingOrCorrupt)?
|
||||
};
|
||||
let part_size = actual_size.max(u64::try_from(part.size).map_err(|_| Error::PartMissingOrCorrupt)?);
|
||||
total.checked_add(part_size).ok_or(Error::PartMissingOrCorrupt)
|
||||
})?
|
||||
};
|
||||
Ok(logical_size.max(persisted_part_size))
|
||||
}
|
||||
|
||||
type UsageVersionPage = StorageListObjectVersionsInfo<ObjectInfo>;
|
||||
|
||||
pub async fn compute_bucket_usage(store: Arc<ECStore>, bucket_name: &str) -> Result<BucketUsageInfo, Error> {
|
||||
@@ -1772,6 +1742,11 @@ pub async fn record_bucket_object_write_unknown_previous_memory(bucket: &str, ne
|
||||
entry.pending_scanner_position = None;
|
||||
}
|
||||
|
||||
/// Fast in-memory increment for immediate quota consistency.
|
||||
pub async fn increment_bucket_usage_memory(bucket: &str, size_increment: u64) {
|
||||
record_bucket_object_write_memory(bucket, None, size_increment).await;
|
||||
}
|
||||
|
||||
/// Fast in-memory update for successful object deletes.
|
||||
pub async fn record_bucket_object_delete_memory(bucket: &str, deleted_size: u64, removed_current_object: bool) {
|
||||
ensure_bucket_usage_cached(bucket).await;
|
||||
@@ -1814,6 +1789,11 @@ pub async fn record_bucket_delete_marker_memory(bucket: &str) {
|
||||
entry.pending_scanner_position = None;
|
||||
}
|
||||
|
||||
/// Fast in-memory decrement for immediate quota consistency
|
||||
pub async fn decrement_bucket_usage_memory(bucket: &str, size_decrement: u64) {
|
||||
record_bucket_object_delete_memory(bucket, size_decrement, size_decrement > 0).await;
|
||||
}
|
||||
|
||||
/// Get bucket usage from the authoritative cache for this topology.
|
||||
async fn get_persisted_bucket_usage(bucket: &str) -> Option<u64> {
|
||||
let store = runtime_sources::object_store_handle()?;
|
||||
@@ -2008,6 +1988,91 @@ pub async fn apply_bucket_usage_memory_overlay(data_usage_info: &mut DataUsageIn
|
||||
apply_bucket_usage_memory_overlay_if_authoritative(data_usage_info, authoritative).await;
|
||||
}
|
||||
|
||||
/// Sync memory cache with backend data (called by scanner)
|
||||
pub async fn sync_memory_cache_with_backend() -> Result<(), Error> {
|
||||
if let Some(store) = runtime_sources::object_store_handle() {
|
||||
match load_data_usage_from_backend(store.clone()).await {
|
||||
Ok(data_usage_info) => {
|
||||
replace_bucket_usage_memory_from_info(&data_usage_info).await;
|
||||
}
|
||||
Err(e) => {
|
||||
debug!("Failed to sync memory cache with backend: {}", e);
|
||||
}
|
||||
}
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Create a data usage cache entry from size summary
|
||||
pub fn create_cache_entry_from_summary(summary: &SizeSummary) -> DataUsageEntry {
|
||||
let mut entry = DataUsageEntry::default();
|
||||
entry.add_sizes(summary);
|
||||
entry
|
||||
}
|
||||
|
||||
/// Convert data usage cache to DataUsageInfo
|
||||
pub fn cache_to_data_usage_info(
|
||||
cache: &DataUsageCache,
|
||||
path: &str,
|
||||
buckets: &[crate::storage_api_contracts::bucket::BucketInfo],
|
||||
) -> DataUsageInfo {
|
||||
let e = match cache.find(path) {
|
||||
Some(e) => e,
|
||||
None => return DataUsageInfo::default(),
|
||||
};
|
||||
let flat = cache.flatten(&e);
|
||||
|
||||
let mut buckets_usage = HashMap::new();
|
||||
for bucket in buckets.iter() {
|
||||
let e = match cache.find(&bucket.name) {
|
||||
Some(e) => e,
|
||||
None => continue,
|
||||
};
|
||||
let flat = cache.flatten(&e);
|
||||
let mut bui = BucketUsageInfo {
|
||||
size: flat.size as u64,
|
||||
versions_count: flat.versions as u64,
|
||||
objects_count: flat.objects as u64,
|
||||
delete_markers_count: flat.delete_markers as u64,
|
||||
object_size_histogram: flat.obj_sizes.to_map(),
|
||||
object_versions_histogram: flat.obj_versions.to_map(),
|
||||
..Default::default()
|
||||
};
|
||||
|
||||
if let Some(rs) = &flat.replication_stats {
|
||||
bui.replica_size = rs.replica_size;
|
||||
bui.replica_count = rs.replica_count;
|
||||
|
||||
for (arn, stat) in rs.targets.iter() {
|
||||
bui.replication_info.insert(
|
||||
arn.clone(),
|
||||
BucketTargetUsageInfo {
|
||||
replication_pending_size: stat.pending_size,
|
||||
replicated_size: stat.replicated_size,
|
||||
replication_failed_size: stat.failed_size,
|
||||
replication_pending_count: stat.pending_count,
|
||||
replication_failed_count: stat.failed_count,
|
||||
replicated_count: stat.replicated_count,
|
||||
..Default::default()
|
||||
},
|
||||
);
|
||||
}
|
||||
}
|
||||
buckets_usage.insert(bucket.name.clone(), bui);
|
||||
}
|
||||
|
||||
DataUsageInfo {
|
||||
last_update: cache.info.last_update,
|
||||
objects_total_count: flat.objects as u64,
|
||||
versions_total_count: flat.versions as u64,
|
||||
delete_markers_total_count: flat.delete_markers as u64,
|
||||
objects_total_size: flat.size as u64,
|
||||
buckets_count: e.children.len() as u64,
|
||||
buckets_usage,
|
||||
..Default::default()
|
||||
}
|
||||
}
|
||||
|
||||
// Helper functions for DataUsageCache operations
|
||||
pub async fn load_data_usage_cache(store: &crate::set_disk::SetDisks, name: &str) -> crate::error::Result<DataUsageCache> {
|
||||
use crate::disk::{BUCKET_META_PREFIX, RUSTFS_META_BUCKET};
|
||||
@@ -3059,102 +3124,6 @@ mod tests {
|
||||
assert_eq!(usage.object_versions_histogram.get("BETWEEN_1000_AND_10000"), Some(&1));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn bucket_usage_uses_the_larger_of_logical_and_physical_size() {
|
||||
let mut metadata = HashMap::new();
|
||||
rustfs_utils::http::insert_str(
|
||||
&mut metadata,
|
||||
rustfs_utils::http::SUFFIX_COMPRESSION,
|
||||
"klauspost/compress/s2".to_string(),
|
||||
);
|
||||
rustfs_utils::http::insert_str(&mut metadata, rustfs_utils::http::SUFFIX_ACTUAL_SIZE, "4096".to_string());
|
||||
let object = ObjectInfo {
|
||||
name: "compressed".to_string(),
|
||||
size: 128,
|
||||
user_defined: Arc::new(metadata),
|
||||
..Default::default()
|
||||
};
|
||||
let mut usage = BucketUsageAccumulator::default();
|
||||
usage
|
||||
.record("bucket", &object)
|
||||
.expect("valid compressed metadata should be counted");
|
||||
assert_eq!(usage.finish().size, 4096);
|
||||
|
||||
let mut framed_metadata = HashMap::new();
|
||||
rustfs_utils::http::insert_str(
|
||||
&mut framed_metadata,
|
||||
rustfs_utils::http::SUFFIX_COMPRESSION,
|
||||
"klauspost/compress/s2".to_string(),
|
||||
);
|
||||
rustfs_utils::http::insert_str(&mut framed_metadata, rustfs_utils::http::SUFFIX_ACTUAL_SIZE, "1".to_string());
|
||||
let framed = ObjectInfo {
|
||||
name: "framed".to_string(),
|
||||
size: 17,
|
||||
user_defined: Arc::new(framed_metadata),
|
||||
..Default::default()
|
||||
};
|
||||
assert_eq!(quota_object_size(&framed).expect("physical framing must remain quota-accounted"), 17);
|
||||
|
||||
let legacy_compressed_part = ObjectInfo {
|
||||
name: "legacy-compressed-part".to_string(),
|
||||
size: 1,
|
||||
user_defined: Arc::new((*framed.user_defined).clone()),
|
||||
parts: Arc::new(vec![rustfs_filemeta::ObjectPartInfo {
|
||||
size: 1,
|
||||
actual_size: -1,
|
||||
..Default::default()
|
||||
}]),
|
||||
..Default::default()
|
||||
};
|
||||
assert_eq!(
|
||||
quota_object_size(&legacy_compressed_part).expect("unknown compressed part size is a valid sentinel"),
|
||||
1
|
||||
);
|
||||
|
||||
let uncompressed_negative_part = ObjectInfo {
|
||||
name: "uncompressed-negative-part".to_string(),
|
||||
size: 1,
|
||||
parts: Arc::new(vec![rustfs_filemeta::ObjectPartInfo {
|
||||
size: 1,
|
||||
actual_size: -1,
|
||||
..Default::default()
|
||||
}]),
|
||||
..Default::default()
|
||||
};
|
||||
assert!(matches!(quota_object_size(&uncompressed_negative_part), Err(Error::PartMissingOrCorrupt)));
|
||||
|
||||
let mut corrupt_metadata = (*object.user_defined).clone();
|
||||
rustfs_utils::http::insert_str(&mut corrupt_metadata, rustfs_utils::http::SUFFIX_ACTUAL_SIZE, "-1".to_string());
|
||||
let corrupt = ObjectInfo {
|
||||
user_defined: Arc::new(corrupt_metadata),
|
||||
..object
|
||||
};
|
||||
assert!(matches!(quota_object_size(&corrupt), Err(Error::PartMissingOrCorrupt)));
|
||||
|
||||
let mut poisoned_metadata = HashMap::new();
|
||||
rustfs_utils::http::insert_str(
|
||||
&mut poisoned_metadata,
|
||||
rustfs_utils::http::SUFFIX_COMPRESSION,
|
||||
"klauspost/compress/s2".to_string(),
|
||||
);
|
||||
rustfs_utils::http::insert_str(&mut poisoned_metadata, rustfs_utils::http::SUFFIX_ACTUAL_SIZE, "1".to_string());
|
||||
let poisoned = ObjectInfo {
|
||||
name: "legacy-swift-metadata".to_string(),
|
||||
size: 4096,
|
||||
user_defined: Arc::new(poisoned_metadata),
|
||||
parts: Arc::new(vec![rustfs_filemeta::ObjectPartInfo {
|
||||
size: 4096,
|
||||
actual_size: 4096,
|
||||
..Default::default()
|
||||
}]),
|
||||
..Default::default()
|
||||
};
|
||||
assert_eq!(
|
||||
quota_object_size(&poisoned).expect("persisted part accounting must bound legacy user metadata"),
|
||||
4096
|
||||
);
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
#[serial]
|
||||
async fn live_bucket_usage_refreshes_are_coalesced_only_while_in_flight() {
|
||||
|
||||
@@ -653,7 +653,6 @@ fn reconcile_servers_with_endpoint_topology(
|
||||
(added, report)
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "exercised by this file's topology tests (backlog#1823)")]
|
||||
fn server_topology_completeness_report(
|
||||
servers: &[ServerProperties],
|
||||
endpoints: &EndpointServerPools,
|
||||
|
||||
@@ -46,49 +46,21 @@ pub(crate) const GET_CODEC_STREAMING_OBJECT_CLASS_MULTIPART: &str = "multipart";
|
||||
pub(crate) const GET_STAGE_DECODE: &str = "decode";
|
||||
pub(crate) const GET_STAGE_EMIT: &str = "emit";
|
||||
pub(crate) const GET_STAGE_FILL: &str = "fill";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_FIRST_BYTE: &str = "first_byte";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_FIRST_METADATA_RESPONSE: &str = "first_metadata_response";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_FIRST_VALID_METADATA_RESPONSE: &str = "first_valid_metadata_response";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_FIRST_SHARD_READ: &str = "first_shard_read";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_FULL_BODY: &str = "full_body";
|
||||
pub(crate) const GET_STAGE_INLINE_PREPARE: &str = "inline_prepare";
|
||||
pub(crate) const GET_STAGE_LOCK_ACQUIRE: &str = "lock_acquire";
|
||||
pub(crate) const GET_STAGE_METADATA: &str = "metadata";
|
||||
pub(crate) const GET_STAGE_METADATA_CACHE_LOOKUP: &str = "metadata_cache_lookup";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_METADATA_FANOUT: &str = "metadata_fanout";
|
||||
pub(crate) const GET_STAGE_METADATA_RESOLVE: &str = "metadata_resolve";
|
||||
pub(crate) const GET_STAGE_OBJECT_INFO: &str = "object_info";
|
||||
pub(crate) const GET_STAGE_OUTPUT_LOCK_WAIT: &str = "output_lock_wait";
|
||||
pub(crate) const GET_STAGE_OUTPUT_POLL: &str = "output_poll";
|
||||
pub(crate) const GET_STAGE_PATH_DECISION: &str = "path_decision";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_QUORUM_REACHED: &str = "quorum_reached";
|
||||
pub(crate) const GET_STAGE_RANGE: &str = "range";
|
||||
pub(crate) const GET_STAGE_READER_SETUP: &str = "reader_setup";
|
||||
@@ -112,28 +84,12 @@ pub(crate) const GET_STAGE_READER_STREAM_FIRST_READ: &str = "reader_stream_first
|
||||
pub(crate) const GET_STAGE_READER_TASK_BITROT_READER_INIT: &str = "reader_task_bitrot_reader_init";
|
||||
pub(crate) const GET_STAGE_READER_TASK_FILE_OPEN: &str = "reader_task_file_open";
|
||||
pub(crate) const GET_STAGE_READER_TASK_READER_CONSTRUCTION: &str = "reader_task_reader_construction";
|
||||
pub(crate) const GET_STAGE_READ_VERSION_DECODE: &str = "read_version_decode";
|
||||
pub(crate) const GET_STAGE_READ_VERSION_PATH_CHECK: &str = "read_version_path_check";
|
||||
pub(crate) const GET_STAGE_READ_VERSION_PATH_RESOLVE: &str = "read_version_path_resolve";
|
||||
pub(crate) const GET_STAGE_READ_VERSION_XLMETA_READ: &str = "read_version_xlmeta_read";
|
||||
pub(crate) const GET_STAGE_RECONSTRUCT: &str = "reconstruct";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_RESPONSE_HANDOFF: &str = "response_handoff";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_SLOWEST_METADATA_RESPONSE: &str = "slowest_metadata_response";
|
||||
pub(crate) const GET_STAGE_STRIPE_READ: &str = "stripe_read";
|
||||
pub(crate) const GET_STAGE_STRIPE_READ_FIRST_SHARD: &str = "stripe_read_first_shard";
|
||||
pub(crate) const GET_STAGE_STRIPE_READ_QUORUM: &str = "stripe_read_quorum";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const GET_STAGE_BITROT_VERIFY: &str = "bitrot_verify";
|
||||
|
||||
pub(crate) const GET_READER_BUFFER_OUTPUT: &str = "output";
|
||||
@@ -181,7 +137,6 @@ pub(crate) const GET_METADATA_CACHE_REASON_NO_LOCK: &str = "no_lock";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_NOT_FOUND_OR_EXPIRED: &str = "not_found_or_expired";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_NOT_READ_DATA: &str = "not_read_data";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_PART_NUMBER: &str = "part_number";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_PART_CHECKSUMS: &str = "part_checksums";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_RAW_DATA_MOVEMENT_READ: &str = "raw_data_movement_read";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_STALE_PUBLICATION: &str = "stale_publication";
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_USABLE: &str = "usable";
|
||||
@@ -190,17 +145,6 @@ pub(crate) const GET_METADATA_CACHE_REASON_VERSION_SUSPENDED: &str = "version_su
|
||||
pub(crate) const GET_METADATA_CACHE_REASON_VERSIONED: &str = "versioned";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_CONFLICTING_METADATA: &str = "conflicting_metadata";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DELETE_MARKER: &str = "delete_marker";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_BODY_VERIFY: &str = "data_read_inline_body_verify";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_DELETED: &str = "data_read_inline_deleted";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_GEOMETRY: &str = "data_read_inline_geometry";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_IDENTITY_MISMATCH: &str = "data_read_inline_identity_mismatch";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_MISSING_PAYLOAD: &str = "data_read_inline_missing_payload";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_MISSING_SHARD: &str = "data_read_inline_missing_shard";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_NOT_INLINE: &str = "data_read_inline_not_inline";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_PART_SHAPE: &str = "data_read_inline_part_shape";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_REMOTE: &str = "data_read_inline_remote";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_SIZE: &str = "data_read_inline_size";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_TRANSFORMED: &str = "data_read_inline_transformed";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_ERROR: &str = "error";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_INSUFFICIENT_QUORUM: &str = "insufficient_quorum";
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_NOT_FOUND: &str = "not_found";
|
||||
@@ -210,20 +154,8 @@ pub(crate) const GET_METADATA_EARLY_STOP_REASON_VERSION_NOT_FOUND: &str = "versi
|
||||
pub(crate) const GET_METADATA_EARLY_STOP_REASON_VERSION_MATCH_QUORUM: &str = "version_match_quorum";
|
||||
|
||||
/// Early-stop active state labels
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const EARLY_STOP_ACTIVE_HIT: &str = "hit";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const EARLY_STOP_ACTIVE_MISS: &str = "miss";
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "GET stage vocabulary; value pinned by this file's tests, no writer yet (backlog#1823)"
|
||||
)]
|
||||
pub(crate) const EARLY_STOP_ACTIVE_DISABLED: &str = "disabled";
|
||||
|
||||
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
|
||||
@@ -509,10 +441,6 @@ mod tests {
|
||||
assert_eq!(GET_STAGE_QUORUM_REACHED, "quorum_reached");
|
||||
assert_eq!(GET_STAGE_RANGE, "range");
|
||||
assert_eq!(GET_STAGE_READER_SETUP, "reader_setup");
|
||||
assert_eq!(GET_STAGE_READ_VERSION_DECODE, "read_version_decode");
|
||||
assert_eq!(GET_STAGE_READ_VERSION_PATH_CHECK, "read_version_path_check");
|
||||
assert_eq!(GET_STAGE_READ_VERSION_PATH_RESOLVE, "read_version_path_resolve");
|
||||
assert_eq!(GET_STAGE_READ_VERSION_XLMETA_READ, "read_version_xlmeta_read");
|
||||
assert_eq!(GET_STAGE_RECONSTRUCT, "reconstruct");
|
||||
assert_eq!(GET_STAGE_RESPONSE_HANDOFF, "response_handoff");
|
||||
assert_eq!(GET_STAGE_SLOWEST_METADATA_RESPONSE, "slowest_metadata_response");
|
||||
@@ -552,7 +480,6 @@ mod tests {
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_NO_LOCK, "no_lock");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_NOT_FOUND_OR_EXPIRED, "not_found_or_expired");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_NOT_READ_DATA, "not_read_data");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_PART_CHECKSUMS, "part_checksums");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_PART_NUMBER, "part_number");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_RAW_DATA_MOVEMENT_READ, "raw_data_movement_read");
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_STALE_PUBLICATION, "stale_publication");
|
||||
@@ -562,32 +489,6 @@ mod tests {
|
||||
assert_eq!(GET_METADATA_CACHE_REASON_VERSIONED, "versioned");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_CONFLICTING_METADATA, "conflicting_metadata");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DELETE_MARKER, "delete_marker");
|
||||
assert_eq!(
|
||||
GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_BODY_VERIFY,
|
||||
"data_read_inline_body_verify"
|
||||
);
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_DELETED, "data_read_inline_deleted");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_GEOMETRY, "data_read_inline_geometry");
|
||||
assert_eq!(
|
||||
GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_IDENTITY_MISMATCH,
|
||||
"data_read_inline_identity_mismatch"
|
||||
);
|
||||
assert_eq!(
|
||||
GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_MISSING_PAYLOAD,
|
||||
"data_read_inline_missing_payload"
|
||||
);
|
||||
assert_eq!(
|
||||
GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_MISSING_SHARD,
|
||||
"data_read_inline_missing_shard"
|
||||
);
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_NOT_INLINE, "data_read_inline_not_inline");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_PART_SHAPE, "data_read_inline_part_shape");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_REMOTE, "data_read_inline_remote");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_SIZE, "data_read_inline_size");
|
||||
assert_eq!(
|
||||
GET_METADATA_EARLY_STOP_REASON_DATA_READ_INLINE_TRANSFORMED,
|
||||
"data_read_inline_transformed"
|
||||
);
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_ERROR, "error");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_INSUFFICIENT_QUORUM, "insufficient_quorum");
|
||||
assert_eq!(GET_METADATA_EARLY_STOP_REASON_NOT_FOUND, "not_found");
|
||||
|
||||
@@ -13,6 +13,8 @@
|
||||
// limitations under the License.
|
||||
|
||||
// #730: diagnostics constants are staged for request-path telemetry migration.
|
||||
#![allow(dead_code)]
|
||||
|
||||
pub(crate) mod admin_server_info;
|
||||
pub(crate) mod get;
|
||||
pub(crate) mod pool;
|
||||
|
||||
@@ -0,0 +1,30 @@
|
||||
// Copyright 2024 RustFS Team
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
//! BytesPool metric label constants.
|
||||
//!
|
||||
//! These constants are used when recording pool acquisition and return
|
||||
//! metrics to avoid string allocations and ensure label consistency.
|
||||
|
||||
/// BytesPool tier labels
|
||||
pub const POOL_TIER_SMALL: &str = "small";
|
||||
pub const POOL_TIER_MEDIUM: &str = "medium";
|
||||
pub const POOL_TIER_LARGE: &str = "large";
|
||||
pub const POOL_TIER_XLARGE: &str = "xlarge";
|
||||
|
||||
/// BytesPool outcome labels
|
||||
pub const POOL_OUTCOME_HIT: &str = "hit";
|
||||
pub const POOL_OUTCOME_MISS: &str = "miss";
|
||||
pub const POOL_OUTCOME_RECYCLED: &str = "recycled";
|
||||
pub const POOL_OUTCOME_DROPPED: &str = "dropped";
|
||||
@@ -241,40 +241,6 @@ pub fn get_drive_list_dir_timeout() -> Duration {
|
||||
)
|
||||
}
|
||||
|
||||
pub(crate) trait DiskStoreRenameDataExt {
|
||||
async fn rename_data_borrowed(
|
||||
&self,
|
||||
src_volume: &str,
|
||||
src_path: &str,
|
||||
fi: &FileInfo,
|
||||
dst_volume: &str,
|
||||
dst_path: &str,
|
||||
) -> Result<RenameDataResp>;
|
||||
}
|
||||
|
||||
impl DiskStoreRenameDataExt for LocalDiskWrapper {
|
||||
async fn rename_data_borrowed(
|
||||
&self,
|
||||
src_volume: &str,
|
||||
src_path: &str,
|
||||
fi: &FileInfo,
|
||||
dst_volume: &str,
|
||||
dst_path: &str,
|
||||
) -> Result<RenameDataResp> {
|
||||
self.track_disk_health_mutation(
|
||||
"rename_data",
|
||||
DiskMetricMutation::Write,
|
||||
|| async {
|
||||
self.disk
|
||||
.rename_data_borrowed(src_volume, src_path, fi, dst_volume, dst_path)
|
||||
.await
|
||||
},
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
}
|
||||
|
||||
pub fn get_drive_walkdir_timeout() -> Duration {
|
||||
get_drive_timeout_duration(
|
||||
rustfs_config::ENV_DRIVE_WALKDIR_TIMEOUT_SECS,
|
||||
@@ -637,23 +603,14 @@ impl Default for DiskOperationMetrics {
|
||||
}
|
||||
|
||||
impl DiskOperationMetrics {
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "internal metrics recorder reached only from record() below (backlog#1823)"
|
||||
)]
|
||||
fn record_call(&mut self) {
|
||||
self.lifetime_calls.fetch_add(1, Ordering::Relaxed);
|
||||
}
|
||||
|
||||
#[allow(
|
||||
dead_code,
|
||||
reason = "internal metrics recorder reached only from record() below (backlog#1823)"
|
||||
)]
|
||||
fn record_latency(&mut self, now_sec: u64, elapsed: Duration) {
|
||||
self.record_latency_atomic(now_sec, elapsed);
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "metrics roll-up with no caller in this port (backlog#1823)")]
|
||||
fn record(&mut self, now_sec: u64, elapsed: Duration) {
|
||||
self.record_call();
|
||||
self.record_latency(now_sec, elapsed);
|
||||
@@ -779,7 +736,6 @@ impl DiskHealthTracker {
|
||||
}
|
||||
|
||||
/// Set disk as faulty
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn set_faulty(&self) {
|
||||
self.status.store(DISK_HEALTH_FAULTY, Ordering::Release);
|
||||
}
|
||||
@@ -860,7 +816,6 @@ impl DiskHealthTracker {
|
||||
became_offline
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn mark_offline(&self, endpoint: &Endpoint, reason: &'static str) -> bool {
|
||||
let current = self.runtime_state();
|
||||
if current == RuntimeDriveHealthState::Offline {
|
||||
@@ -991,13 +946,11 @@ impl DiskHealthTracker {
|
||||
}
|
||||
|
||||
/// Get waiting operations count
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn waiting_count(&self) -> u32 {
|
||||
self.waiting.load(Ordering::Relaxed)
|
||||
}
|
||||
|
||||
/// Get last success timestamp
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn last_success(&self) -> i64 {
|
||||
self.last_success.load(Ordering::Acquire)
|
||||
}
|
||||
@@ -1039,6 +992,21 @@ impl Default for DiskHealthTracker {
|
||||
}
|
||||
}
|
||||
|
||||
/// Health check context key for tracking disk operations
|
||||
#[derive(Debug, Clone)]
|
||||
struct HealthDiskCtxKey;
|
||||
|
||||
#[derive(Debug)]
|
||||
struct HealthDiskCtxValue {
|
||||
last_success: Arc<AtomicI64>,
|
||||
}
|
||||
|
||||
impl HealthDiskCtxValue {
|
||||
fn log_success(&self) {
|
||||
self.last_success.store(current_unix_nanos(), Ordering::Relaxed);
|
||||
}
|
||||
}
|
||||
|
||||
/// LocalDiskWrapper wraps a DiskStore with health tracking capabilities.
|
||||
/// This is similar to Go's xlStorageDiskIDCheck.
|
||||
#[derive(Debug, Clone)]
|
||||
@@ -1070,6 +1038,10 @@ impl LocalDiskWrapper {
|
||||
)
|
||||
}
|
||||
|
||||
pub(crate) fn new_with_health(disk: Arc<LocalDisk>, health_check: bool, health: Arc<DiskHealthTracker>) -> Self {
|
||||
Self::new_with_health_and_metrics(disk, health_check, health, Arc::new(DiskHealthMetricEpoch::default()))
|
||||
}
|
||||
|
||||
pub(crate) fn new_with_reconnect_state(
|
||||
disk: Arc<LocalDisk>,
|
||||
health_check: bool,
|
||||
@@ -1432,6 +1404,20 @@ impl LocalDiskWrapper {
|
||||
}
|
||||
}
|
||||
|
||||
async fn check_id(&self, want_id: Option<Uuid>) -> Result<()> {
|
||||
if want_id.is_none() {
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
let stored_disk_id = self.disk.get_disk_id().await?;
|
||||
|
||||
if stored_disk_id != want_id {
|
||||
return Err(Error::other(format!("Disk ID mismatch wanted {want_id:?}, got {stored_disk_id:?}")));
|
||||
}
|
||||
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Check if disk ID is stale
|
||||
async fn check_disk_stale(&self) -> Result<()> {
|
||||
let Some(current_disk_id) = *self.disk_id.read().await else {
|
||||
@@ -1999,8 +1985,13 @@ impl DiskAPI for LocalDiskWrapper {
|
||||
dst_volume: &str,
|
||||
dst_path: &str,
|
||||
) -> Result<RenameDataResp> {
|
||||
self.rename_data_borrowed(src_volume, src_path, &fi, dst_volume, dst_path)
|
||||
.await
|
||||
self.track_disk_health_mutation(
|
||||
"rename_data",
|
||||
DiskMetricMutation::Write,
|
||||
|| async { self.disk.rename_data(src_volume, src_path, fi, dst_volume, dst_path).await },
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
|
||||
async fn list_dir(&self, origvolume: &str, volume: &str, dir_path: &str, count: i32) -> Result<Vec<String>> {
|
||||
@@ -2031,21 +2022,6 @@ impl DiskAPI for LocalDiskWrapper {
|
||||
.await
|
||||
}
|
||||
|
||||
async fn read_file_stream_chunks(
|
||||
&self,
|
||||
volume: &str,
|
||||
path: &str,
|
||||
offset: usize,
|
||||
length: usize,
|
||||
) -> Result<Option<rustfs_rio::ChunkReaderBox>> {
|
||||
self.track_disk_health_with_op(
|
||||
"read_file_stream_chunks",
|
||||
|| async { self.disk.read_file_stream_chunks(volume, path, offset, length).await },
|
||||
get_max_timeout_duration(),
|
||||
)
|
||||
.await
|
||||
}
|
||||
|
||||
async fn read_file_mmap_copy(&self, volume: &str, path: &str, offset: usize, length: usize) -> Result<bytes::Bytes> {
|
||||
self.track_disk_health_with_op(
|
||||
"read_file_mmap_copy",
|
||||
|
||||
@@ -113,9 +113,6 @@ pub enum DiskError {
|
||||
#[error("bit-rot hash algorithm is invalid")]
|
||||
BitrotHashAlgoInvalid,
|
||||
|
||||
/// Never constructed locally by RustFS (only reachable through wire
|
||||
/// decoding, and no current node sends it). The wire code is kept for
|
||||
/// cross-version compatibility — do not renumber or remove (backlog#1831).
|
||||
#[error("Rename across devices not allowed, please fix your backend configuration")]
|
||||
CrossDeviceLink,
|
||||
|
||||
@@ -146,9 +143,6 @@ pub enum DiskError {
|
||||
#[error("io error {0}")]
|
||||
Io(#[source] io::Error),
|
||||
|
||||
/// Never constructed locally by RustFS (only reachable through wire
|
||||
/// decoding, and no current node sends it). The wire code is kept for
|
||||
/// cross-version compatibility — do not renumber or remove (backlog#1831).
|
||||
#[error("source stalled")]
|
||||
SourceStalled,
|
||||
|
||||
@@ -648,6 +642,19 @@ impl Hash for DiskError {
|
||||
// is currently commented out to avoid complexity. These can be re-enabled
|
||||
// when needed for specific disk quorum checking and error aggregation logic.
|
||||
|
||||
/// Bitrot errors
|
||||
#[derive(Debug, thiserror::Error)]
|
||||
pub enum BitrotErrorType {
|
||||
#[error("bitrot checksum verification failed")]
|
||||
BitrotChecksumMismatch { expected: String, got: String },
|
||||
}
|
||||
|
||||
impl From<BitrotErrorType> for DiskError {
|
||||
fn from(e: BitrotErrorType) -> Self {
|
||||
DiskError::other(e)
|
||||
}
|
||||
}
|
||||
|
||||
/// Context wrapper for file access errors
|
||||
#[derive(Debug, thiserror::Error)]
|
||||
pub struct FileAccessDeniedWithContext {
|
||||
@@ -862,6 +869,19 @@ mod tests {
|
||||
let _disk_error: DiskError = json_error.into();
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_bitrot_error_type() {
|
||||
let bitrot_error = BitrotErrorType::BitrotChecksumMismatch {
|
||||
expected: "abc123".to_string(),
|
||||
got: "def456".to_string(),
|
||||
};
|
||||
|
||||
assert!(bitrot_error.to_string().contains("bitrot checksum verification failed"));
|
||||
|
||||
let disk_error: DiskError = bitrot_error.into();
|
||||
assert!(matches!(disk_error, DiskError::Io(_)));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_file_access_denied_with_context() {
|
||||
let path = PathBuf::from("/test/path");
|
||||
|
||||
@@ -48,7 +48,6 @@ pub fn to_volume_error(io_err: std::io::Error) -> std::io::Error {
|
||||
}
|
||||
}
|
||||
|
||||
#[allow(dead_code, reason = "asserted by this file's tests (backlog#1823)")]
|
||||
pub fn to_disk_error(io_err: std::io::Error) -> std::io::Error {
|
||||
match io_err.kind() {
|
||||
std::io::ErrorKind::NotFound => DiskError::DiskNotFound.into(),
|
||||
|
||||
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user