fix(get): validate fast path body before response commit (#2495)

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: houseme <4829346+houseme@users.noreply.github.com>
This commit is contained in:
houseme
2026-04-12 01:04:22 +08:00
committed by GitHub
parent c8c71e34b6
commit eb0dc24921
4 changed files with 308 additions and 219 deletions
+6
View File
@@ -25,6 +25,12 @@ Current `test-and-lint` gate includes:
- `cargo nextest run --all --exclude e2e_test`
- `cargo test --all --doc`
- `cargo test -p rustfs get_object_chunk_fast_path`
- `cargo test -p rustfs materialize_chunk_stream_before_commit`
- `touch rustfs/build.rs`
- `cargo build -p rustfs --bins --jobs 2`
- `cargo test -p e2e_test archive_multipart_roundtrip_preserves_bytes`
- `cargo test -p e2e_test presigned_get_and_reverse_proxy_preserve_multipart_bytes_with_fast_path`
- `cargo fmt --all --check`
- `cargo clippy --all-targets --all-features -- -D warnings`
- `./scripts/check_layer_dependencies.sh`
Generated
+64 -50
View File
@@ -56,13 +56,13 @@ dependencies = [
[[package]]
name = "aes"
version = "0.9.0-rc.4"
version = "0.9.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "04097e08a47d9ad181c2e1f4a5fabc9ae06ce8839a333ba9a949bcb0d31fd2a3"
checksum = "66bd29a732b644c0431c6140f370d097879203d79b80c94a6747ba0872adaef8"
dependencies = [
"cipher 0.5.1",
"cpubits",
"cpufeatures 0.2.17",
"cpufeatures 0.3.0",
]
[[package]]
@@ -72,7 +72,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e22c0c90bbe8d4f77c3ca9ddabe41a1f8382d6fc1f7cea89459d0f320371f972"
dependencies = [
"aead",
"aes 0.9.0-rc.4",
"aes 0.9.0",
"cipher 0.5.1",
"ctr",
"ghash",
@@ -410,7 +410,7 @@ dependencies = [
"arrow-schema",
"chrono",
"half",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itoa",
"lexical-core",
"memchr",
@@ -1425,9 +1425,9 @@ checksum = "37b2a672a2cb129a2e41c10b1224bb368f9f37a2b16b612598138befd7b37eb5"
[[package]]
name = "cc"
version = "1.2.59"
version = "1.2.60"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b7a4d3ec6524d28a329fc53654bbadc9bdd7b0431f5d65f1a56ffb28a1ee5283"
checksum = "43c5703da9466b66a946814e1adf53ea2c90f10063b86290cc9eb67ce3478a20"
dependencies = [
"find-msvc-tools",
"jobserver",
@@ -2345,7 +2345,7 @@ dependencies = [
"chrono",
"half",
"hashbrown 0.16.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"libc",
"log",
@@ -2549,7 +2549,7 @@ dependencies = [
"datafusion-functions-aggregate-common",
"datafusion-functions-window-common",
"datafusion-physical-expr-common",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"paste",
"recursive",
@@ -2565,7 +2565,7 @@ checksum = "ab05fdd00e05d5a6ee362882546d29d6d3df43a6c55355164a7fbee12d163bc9"
dependencies = [
"arrow",
"datafusion-common",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"paste",
]
@@ -2729,7 +2729,7 @@ dependencies = [
"datafusion-expr",
"datafusion-expr-common",
"datafusion-physical-expr",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"log",
"recursive",
@@ -2752,7 +2752,7 @@ dependencies = [
"datafusion-physical-expr-common",
"half",
"hashbrown 0.16.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"parking_lot 0.12.5",
"paste",
@@ -2788,7 +2788,7 @@ dependencies = [
"datafusion-common",
"datafusion-expr-common",
"hashbrown 0.16.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"parking_lot 0.12.5",
]
@@ -2835,7 +2835,7 @@ dependencies = [
"futures",
"half",
"hashbrown 0.16.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itertools 0.14.0",
"log",
"num-traits",
@@ -2887,7 +2887,7 @@ dependencies = [
"datafusion-common",
"datafusion-expr",
"datafusion-functions-nested",
"indexmap 2.13.1",
"indexmap 2.14.0",
"log",
"recursive",
"regex",
@@ -3065,7 +3065,7 @@ dependencies = [
[[package]]
name = "dial9-tokio-telemetry"
version = "0.2.0"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=60502082601b647c4a51962595721f631b7bbce1#60502082601b647c4a51962595721f631b7bbce1"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=2117416f6052fb66edb393f2a8a22c1fdcb91bbd#2117416f6052fb66edb393f2a8a22c1fdcb91bbd"
dependencies = [
"arc-swap",
"bon",
@@ -3089,7 +3089,7 @@ dependencies = [
[[package]]
name = "dial9-trace-format"
version = "0.2.0"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=60502082601b647c4a51962595721f631b7bbce1#60502082601b647c4a51962595721f631b7bbce1"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=2117416f6052fb66edb393f2a8a22c1fdcb91bbd#2117416f6052fb66edb393f2a8a22c1fdcb91bbd"
dependencies = [
"dial9-trace-format-derive",
"serde",
@@ -3098,7 +3098,7 @@ dependencies = [
[[package]]
name = "dial9-trace-format-derive"
version = "0.2.0"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=60502082601b647c4a51962595721f631b7bbce1#60502082601b647c4a51962595721f631b7bbce1"
source = "git+https://github.com/dial9-rs/dial9-tokio-telemetry.git?rev=2117416f6052fb66edb393f2a8a22c1fdcb91bbd#2117416f6052fb66edb393f2a8a22c1fdcb91bbd"
dependencies = [
"proc-macro2",
"quote",
@@ -4072,7 +4072,7 @@ dependencies = [
"futures-core",
"futures-sink",
"http 1.4.0",
"indexmap 2.13.1",
"indexmap 2.14.0",
"slab",
"tokio",
"tokio-util",
@@ -4138,6 +4138,20 @@ dependencies = [
"serde_core",
]
[[package]]
name = "hashbrown"
version = "0.17.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4f467dd6dccf739c208452f8014c75c18bb8301b050ad1cfb27153803edb0f51"
dependencies = [
"allocator-api2",
"equivalent",
"foldhash 0.2.0",
"rayon",
"serde",
"serde_core",
]
[[package]]
name = "headers"
version = "0.4.1"
@@ -4572,12 +4586,12 @@ dependencies = [
[[package]]
name = "indexmap"
version = "2.13.1"
version = "2.14.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "45a8a2b9cb3e0b0c1803dbb0758ffac5de2f425b23c28f518faabd9d805342ff"
checksum = "d466e9454f08e4a911e14806c24e16fba1b4c121d1ea474396f396069cf949d9"
dependencies = [
"equivalent",
"hashbrown 0.16.1",
"hashbrown 0.17.0",
"serde",
"serde_core",
]
@@ -4589,7 +4603,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "232929e1d75fe899576a3d5c7416ad0d88dbfbb3c3d6aa00873a7408a50ddb88"
dependencies = [
"ahash 0.8.12",
"indexmap 2.13.1",
"indexmap 2.14.0",
"is-terminal",
"itoa",
"log",
@@ -4612,7 +4626,7 @@ dependencies = [
"crossbeam-utils",
"dashmap",
"env_logger",
"indexmap 2.13.1",
"indexmap 2.14.0",
"itoa",
"log",
"num-format",
@@ -5056,14 +5070,14 @@ dependencies = [
[[package]]
name = "libredox"
version = "0.1.15"
version = "0.1.16"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7ddbf48fd451246b1f8c2610bd3b4ac0cc6e149d89832867093ab69a17194f08"
checksum = "e02f3bb43d335493c96bf3fd3a321600bf6bd07ed34bc64118e9293bdffea46c"
dependencies = [
"bitflags 2.11.0",
"libc",
"plain",
"redox_syscall 0.7.3",
"redox_syscall 0.7.4",
]
[[package]]
@@ -5345,7 +5359,7 @@ dependencies = [
"crossbeam-epoch",
"crossbeam-utils",
"hashbrown 0.16.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"metrics",
"ordered-float 5.3.0",
"quanta",
@@ -6371,7 +6385,7 @@ checksum = "8701b58ea97060d5e5b155d383a69952a60943f0e6dfe30b04c287beb0b27455"
dependencies = [
"fixedbitset",
"hashbrown 0.15.5",
"indexmap 2.13.1",
"indexmap 2.14.0",
"serde",
]
@@ -6801,7 +6815,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b4aeaa1f2460f1d348eeaeed86aea999ce98c1bded6f089ff8514c9d9dbdc973"
dependencies = [
"anyhow",
"indexmap 2.13.1",
"indexmap 2.14.0",
"log",
"protobuf",
"protobuf-support",
@@ -7198,9 +7212,9 @@ dependencies = [
[[package]]
name = "redox_syscall"
version = "0.7.3"
version = "0.7.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6ce70a74e890531977d37e532c34d45e9055d2409ed08ddba14529471ed0be16"
checksum = "f450ad9c3b1da563fb6948a8e0fb0fb9269711c9c73d9ea1de5058c79c8d643a"
dependencies = [
"bitflags 2.11.0",
]
@@ -7448,9 +7462,9 @@ checksum = "3582f63211428f83597b51b2ddb88e2a91a9d52d12831f9d08f5e624e8977422"
[[package]]
name = "rmcp"
version = "1.3.0"
version = "1.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2231b2c085b371c01bc90c0e6c1cab8834711b6394533375bdbf870b0166d419"
checksum = "f542f74cf247da16f19bbc87e298cd201e912314f4083e88cdd671f44f5fcb53"
dependencies = [
"async-trait",
"base64 0.22.1",
@@ -7470,9 +7484,9 @@ dependencies = [
[[package]]
name = "rmcp-macros"
version = "1.3.0"
version = "1.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "36ea0e100fadf81be85d7ff70f86cd805c7572601d4ab2946207f36540854b43"
checksum = "b2391e4ae47f314e70eaafb6c7bd82e495e770b935448864446302143019151f"
dependencies = [
"darling 0.23.0",
"proc-macro2",
@@ -7669,7 +7683,7 @@ dependencies = [
"flatbuffers",
"futures",
"futures-util",
"hashbrown 0.16.1",
"hashbrown 0.17.0",
"hex-simd",
"http 1.4.0",
"http-body 1.0.1",
@@ -7780,7 +7794,7 @@ dependencies = [
"chrono",
"const-str",
"futures",
"hashbrown 0.16.1",
"hashbrown 0.17.0",
"metrics",
"rumqttc-next",
"rustfs-config",
@@ -8207,7 +8221,7 @@ dependencies = [
"chrono",
"form_urlencoded",
"futures",
"hashbrown 0.16.1",
"hashbrown 0.17.0",
"quick-xml 0.39.2",
"rayon",
"rumqttc-next",
@@ -8586,7 +8600,7 @@ dependencies = [
"crc-fast",
"flate2",
"futures",
"hashbrown 0.16.1",
"hashbrown 0.17.0",
"hex-simd",
"highway",
"hmac 0.13.0",
@@ -8775,9 +8789,9 @@ checksum = "f87165f0995f63a9fbeea62b64d10b4d9d8e78ec6d7d51fb2125fda7bb36788f"
[[package]]
name = "rustls-webpki"
version = "0.103.10"
version = "0.103.11"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "df33b2b81ac578cabaf06b89b0631153a3f416b0a886e8a7a1707fb51abbd1ef"
checksum = "20a6af516fea4b20eccceaf166e8aa666ac996208e8a644ce3ef5aa783bc7cd4"
dependencies = [
"aws-lc-rs",
"ring",
@@ -9098,7 +9112,7 @@ dependencies = [
"chrono",
"hex",
"indexmap 1.9.3",
"indexmap 2.13.1",
"indexmap 2.14.0",
"schemars 0.9.0",
"schemars 1.2.1",
"serde_core",
@@ -10118,7 +10132,7 @@ checksum = "ebe5ef63511595f1344e2d5cfa636d973292adc0eec1f0ad45fae9f0851ab1d4"
dependencies = [
"futures-core",
"futures-util",
"indexmap 2.13.1",
"indexmap 2.14.0",
"pin-project-lite",
"slab",
"sync_wrapper",
@@ -10621,7 +10635,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "bb0e353e6a2fbdc176932bbaab493762eb1255a7900fe0fea1a2f96c296cc909"
dependencies = [
"anyhow",
"indexmap 2.13.1",
"indexmap 2.14.0",
"wasm-encoder",
"wasmparser",
]
@@ -10647,7 +10661,7 @@ checksum = "47b807c72e1bac69382b3a6fb3dbe8ea4c0ed87ff5629b8685ae6b9a611028fe"
dependencies = [
"bitflags 2.11.0",
"hashbrown 0.15.5",
"indexmap 2.13.1",
"indexmap 2.14.0",
"semver",
]
@@ -11121,7 +11135,7 @@ checksum = "b7c566e0f4b284dd6561c786d9cb0142da491f46a9fbed79ea69cdad5db17f21"
dependencies = [
"anyhow",
"heck",
"indexmap 2.13.1",
"indexmap 2.14.0",
"prettyplease",
"syn 2.0.117",
"wasm-metadata",
@@ -11152,7 +11166,7 @@ checksum = "9d66ea20e9553b30172b5e831994e35fbde2d165325bec84fc43dbf6f4eb9cb2"
dependencies = [
"anyhow",
"bitflags 2.11.0",
"indexmap 2.13.1",
"indexmap 2.14.0",
"log",
"serde",
"serde_derive",
@@ -11171,7 +11185,7 @@ checksum = "ecc8ac4bc1dc3381b7f59c34f00b67e18f910c2c0f50015669dde7def656a736"
dependencies = [
"anyhow",
"id-arena",
"indexmap 2.13.1",
"indexmap 2.14.0",
"log",
"semver",
"serde",
@@ -11409,7 +11423,7 @@ dependencies = [
"flate2",
"getrandom 0.4.2",
"hmac 0.12.1",
"indexmap 2.13.1",
"indexmap 2.14.0",
"lzma-rust2",
"memchr",
"pbkdf2 0.12.2",
+3 -3
View File
@@ -154,7 +154,7 @@ flatbuffers = "25.12.19"
form_urlencoded = "1.2.2"
prost = "0.14.3"
quick-xml = "0.39.2"
rmcp = { version = "1.3.0" }
rmcp = { version = "1.4.0" }
rmp = { version = "0.8.15" }
rmp-serde = { version = "1.3.1" }
serde = { version = "1.0.228", features = ["derive"] }
@@ -218,7 +218,7 @@ flate2 = "1.1.9"
glob = "0.3.3"
google-cloud-storage = "1.10.0"
google-cloud-auth = "1.8.0"
hashbrown = { version = "0.16.1", features = ["serde", "rayon"] }
hashbrown = { version = "0.17.0", features = ["serde", "rayon"] }
hex = "0.4.3"
hex-simd = "0.8.0"
highway = { version = "1.3.0" }
@@ -288,7 +288,7 @@ zstd = "0.13.3"
# Observability and Metrics
metrics = "0.24.3"
dial9-tokio-telemetry = { version = "0.2", git = "https://github.com/dial9-rs/dial9-tokio-telemetry.git", rev = "60502082601b647c4a51962595721f631b7bbce1" }
dial9-tokio-telemetry = { version = "0.2", git = "https://github.com/dial9-rs/dial9-tokio-telemetry.git", rev = "2117416f6052fb66edb393f2a8a22c1fdcb91bbd" }
opentelemetry = { version = "0.31.0" }
opentelemetry-appender-tracing = { version = "0.31.1", features = ["experimental_use_tracing_span_context", "experimental_metadata_attributes", "spec_unstable_logs_enabled"] }
opentelemetry-otlp = { version = "0.31.1", features = ["gzip-http", "reqwest-rustls"] }
+235 -166
View File
@@ -14,25 +14,28 @@
use super::DeadlockRequestGuard;
use super::GetObjectRequestContext;
use super::get_object_zero_copy::{GetObjectIoPlanning, GetObjectPreparedRead, prepare_get_object_read_execution};
use super::get_object_zero_copy::{
GetObjectIoPlanning, GetObjectPreparedRead, prepare_get_object_read, prepare_get_object_read_execution,
};
use crate::error::ApiError;
use crate::storage::concurrency::{ConcurrencyManager, GetObjectGuard, get_buffer_size_opt_in};
use crate::storage::get_validated_store;
use crate::storage::options::filter_object_metadata;
use crate::storage::timeout_wrapper::{RequestTimeoutWrapper, TimeoutConfig};
use bytes::Bytes;
use futures_util::StreamExt;
use rustfs_ecstore::bucket::versioning_sys::BucketVersioningSys;
use rustfs_ecstore::error::StorageError;
use rustfs_ecstore::store_api::{GetObjectChunkPath, HTTPRangeSpec, ObjectInfo};
use rustfs_ecstore::store_api::{HTTPRangeSpec, ObjectInfo};
use rustfs_io_core::BoxChunkStream;
use rustfs_object_io::get::{
GetObjectBodyPlan as ObjectIoGetObjectBodyPlan, GetObjectBodyPlanningInputs as ObjectIoGetObjectBodyPlanningInputs,
GetObjectBodySource, GetObjectDataPlaneMetricContract as ObjectIoGetObjectDataPlaneMetricContract, GetObjectFlowResult,
GetObjectOutputContext, GetObjectReadSetup, MaterializeGetObjectBodyError as ObjectIoMaterializeGetObjectBodyError,
build_chunk_blob as object_io_build_chunk_blob,
build_cors_wrapped_get_object_flow_result as object_io_build_cors_wrapped_get_object_flow_result,
build_get_object_checksums as object_io_build_get_object_checksums,
build_get_object_output_context as object_io_build_get_object_output_context,
chunk_body_data_plane_labels as object_io_chunk_body_data_plane_labels,
build_memory_blob as object_io_build_memory_blob, chunk_body_data_plane_labels as object_io_chunk_body_data_plane_labels,
get_object_chunk_path_label as object_io_get_object_chunk_path_label,
materialize_get_object_body as object_io_materialize_get_object_body, plan_get_object_body as object_io_plan_get_object_body,
plan_get_object_strategy_layout as object_io_plan_get_object_strategy_layout,
@@ -51,63 +54,93 @@ pub(super) struct GetObjectBootstrap {
pub(super) _deadlock_request_guard: DeadlockRequestGuard,
}
fn classify_get_object_midstream_error(err: &std::io::Error) -> &'static str {
let lower = err.to_string().to_ascii_lowercase();
if lower.contains("bitrot") {
"bitrot"
} else if lower.contains("decode") {
"decode"
} else {
match err.kind() {
std::io::ErrorKind::UnexpectedEof => "unexpected_eof",
std::io::ErrorKind::BrokenPipe | std::io::ErrorKind::ConnectionAborted => "channel_closed",
_ => "io_other",
}
}
#[derive(Debug)]
struct ChunkCommitMaterializationError {
source: std::io::Error,
streamed_bytes: usize,
}
fn instrument_get_object_chunk_stream(
request_context: &GetObjectRequestContext,
chunk_stream: BoxChunkStream,
path: GetObjectChunkPath,
copy_mode: rustfs_io_metrics::CopyMode,
response_content_length: i64,
) -> BoxChunkStream {
let bucket = request_context.bucket.clone();
let key = request_context.key.clone();
let version_id = request_context.opts.version_id.clone();
let path_label = object_io_get_object_chunk_path_label(path);
let mut sent_bytes = 0usize;
fn build_chunk_materialization_length_error(actual: usize, expected: usize) -> std::io::Error {
let error_kind = if actual > expected {
std::io::ErrorKind::InvalidData
} else {
std::io::ErrorKind::UnexpectedEof
};
Box::pin(chunk_stream.map(move |result| match result {
Ok(chunk) => {
sent_bytes = sent_bytes.saturating_add(chunk.len());
Ok(chunk)
std::io::Error::new(
error_kind,
format!("chunk fast path produced {actual} bytes before response commit, expected {expected}"),
)
}
async fn materialize_chunk_stream_before_commit_with_threshold(
mut chunk_stream: BoxChunkStream,
response_content_length: i64,
optimal_buffer_size: usize,
in_memory_threshold_bytes: usize,
) -> Result<Option<StreamingBlob>, ChunkCommitMaterializationError> {
let expected_bytes = usize::try_from(response_content_length).map_err(|_| ChunkCommitMaterializationError {
source: std::io::Error::new(
std::io::ErrorKind::InvalidInput,
format!("negative response content length {response_content_length} for chunk fast path"),
),
streamed_bytes: 0,
})?;
// Objects larger than the in-memory threshold fall back to the legacy reader path
// rather than spooling to disk, to avoid exhausting local disk under concurrent large downloads.
if expected_bytes > in_memory_threshold_bytes {
return Err(ChunkCommitMaterializationError {
source: std::io::Error::other(format!(
"chunk fast path object size {expected_bytes} exceeds in-memory threshold \
{in_memory_threshold_bytes}; falling back to legacy reader"
)),
streamed_bytes: 0,
});
}
let mut buf = Vec::with_capacity(expected_bytes);
let mut streamed_bytes = 0usize;
while let Some(result) = chunk_stream.next().await {
let chunk = result.map_err(|source| ChunkCommitMaterializationError { source, streamed_bytes })?;
let bytes = chunk.as_bytes();
streamed_bytes = streamed_bytes.saturating_add(bytes.len());
if streamed_bytes > expected_bytes {
return Err(ChunkCommitMaterializationError {
source: build_chunk_materialization_length_error(streamed_bytes, expected_bytes),
streamed_bytes,
});
}
Err(err) => {
let error_kind = classify_get_object_midstream_error(&err);
rustfs_io_metrics::record_get_object_fast_path_midstream_error(
path_label,
copy_mode,
error_kind,
sent_bytes,
response_content_length,
);
warn!(
bucket = %bucket,
key = %key,
version_id = ?version_id,
path = path_label,
copy_mode = copy_mode.as_str(),
promised_bytes = response_content_length,
sent_bytes,
error_kind,
error = %err,
"GetObject chunk fast path failed mid-stream after response commit"
);
Err(err)
}
}))
buf.extend_from_slice(bytes.as_ref());
}
if streamed_bytes != expected_bytes {
return Err(ChunkCommitMaterializationError {
source: build_chunk_materialization_length_error(streamed_bytes, expected_bytes),
streamed_bytes,
});
}
Ok(object_io_build_memory_blob(
Bytes::from(buf),
response_content_length,
optimal_buffer_size,
))
}
async fn materialize_chunk_stream_before_commit(
chunk_stream: BoxChunkStream,
response_content_length: i64,
optimal_buffer_size: usize,
) -> Result<Option<StreamingBlob>, ChunkCommitMaterializationError> {
materialize_chunk_stream_before_commit_with_threshold(
chunk_stream,
response_content_length,
optimal_buffer_size,
rustfs_config::DEFAULT_OBJECT_SEEK_SUPPORT_THRESHOLD,
)
.await
}
async fn build_get_object_body_adapter<R>(
@@ -307,79 +340,14 @@ pub(super) async fn build_get_object_output_context(
let bucket = &request_context.bucket;
let key = &request_context.key;
let part_number = request_context.part_number;
let GetObjectReadSetup {
info,
event_info,
body_source,
rs,
content_type,
last_modified,
response_content_length,
content_range,
server_side_encryption,
sse_customer_algorithm,
sse_customer_key_md5,
ssekms_key_id,
encryption_applied,
} = read_setup;
let mut active_read_setup = read_setup;
let optimal_buffer_size = finalize_get_object_strategy_runtime(
request_context,
rs.as_ref(),
manager,
base_buffer_size,
&info,
response_content_length,
io_planning,
);
let (body, metric_contract) = match body_source {
GetObjectBodySource::Reader(final_stream) => {
let body = build_get_object_body_adapter(
final_stream,
bucket,
key,
response_content_length,
optimal_buffer_size,
ObjectIoGetObjectBodyPlanningInputs {
is_part_request: part_number.is_some(),
is_range_request: rs.is_some(),
encryption_applied,
response_size: response_content_length,
},
)
.await?;
let metric_contract = ObjectIoGetObjectDataPlaneMetricContract::disk(
rustfs_io_metrics::IoPath::Legacy,
rustfs_io_metrics::CopyMode::SingleCopy,
);
(body, metric_contract)
}
GetObjectBodySource::Chunk {
stream: chunk_stream,
path,
copy_mode,
} => {
let (io_path, copy_mode) = object_io_chunk_body_data_plane_labels(path, copy_mode);
let chunk_stream =
instrument_get_object_chunk_stream(request_context, chunk_stream, path, copy_mode, response_content_length);
(
object_io_build_chunk_blob(chunk_stream),
ObjectIoGetObjectDataPlaneMetricContract::disk(io_path, copy_mode),
)
}
};
let checksums = object_io_build_get_object_checksums(&info, &request_context.headers, part_number, rs.as_ref())
.map_err(ApiError::from)?;
let filtered_metadata = filter_object_metadata(&info.user_defined);
Ok((
object_io_build_get_object_output_context(
body,
loop {
let GetObjectReadSetup {
info,
event_info,
body_source,
rs,
content_type,
last_modified,
response_content_length,
@@ -388,14 +356,108 @@ pub(super) async fn build_get_object_output_context(
sse_customer_algorithm,
sse_customer_key_md5,
ssekms_key_id,
&checksums,
filtered_metadata,
versioned,
optimal_buffer_size,
Some(metric_contract.copy_mode),
),
metric_contract,
))
encryption_applied,
} = active_read_setup;
let optimal_buffer_size = finalize_get_object_strategy_runtime(
request_context,
rs.as_ref(),
manager,
base_buffer_size,
&info,
response_content_length,
io_planning,
);
let (body, metric_contract) = match body_source {
GetObjectBodySource::Reader(final_stream) => {
let body = build_get_object_body_adapter(
final_stream,
bucket,
key,
response_content_length,
optimal_buffer_size,
ObjectIoGetObjectBodyPlanningInputs {
is_part_request: part_number.is_some(),
is_range_request: rs.is_some(),
encryption_applied,
response_size: response_content_length,
},
)
.await?;
let metric_contract = ObjectIoGetObjectDataPlaneMetricContract::disk(
rustfs_io_metrics::IoPath::Legacy,
rustfs_io_metrics::CopyMode::SingleCopy,
);
(body, metric_contract)
}
GetObjectBodySource::Chunk {
stream: chunk_stream,
path,
copy_mode,
} => {
let (io_path, copy_mode) = object_io_chunk_body_data_plane_labels(path, copy_mode);
match materialize_chunk_stream_before_commit(chunk_stream, response_content_length, optimal_buffer_size).await {
Ok(body) => (body, ObjectIoGetObjectDataPlaneMetricContract::disk(io_path, copy_mode)),
Err(err) => {
let path_label = object_io_get_object_chunk_path_label(path);
rustfs_io_metrics::record_io_fallback(
rustfs_io_metrics::IoStage::ReadSetup,
rustfs_io_metrics::FallbackReason::ProbeFailed,
);
rustfs_io_metrics::record_get_object_fast_path_probe_failed(
path_label,
copy_mode,
response_content_length,
);
warn!(
bucket = %request_context.bucket,
key = %request_context.key,
version_id = ?request_context.opts.version_id,
path = path_label,
copy_mode = copy_mode.as_str(),
promised_bytes = response_content_length,
materialized_bytes = err.streamed_bytes,
error = %err.source,
"GetObject chunk fast path full-body materialization failed before response commit"
);
let store = get_validated_store(&request_context.bucket).await?;
active_read_setup =
prepare_get_object_read(request_context, &store, manager, std::time::Instant::now()).await?;
continue;
}
}
}
};
let checksums = object_io_build_get_object_checksums(&info, &request_context.headers, part_number, rs.as_ref())
.map_err(ApiError::from)?;
let filtered_metadata = filter_object_metadata(&info.user_defined);
return Ok((
object_io_build_get_object_output_context(
body,
info,
event_info,
content_type,
last_modified,
response_content_length,
content_range,
server_side_encryption,
sse_customer_algorithm,
sse_customer_key_md5,
ssekms_key_id,
&checksums,
filtered_metadata,
versioned,
optimal_buffer_size,
Some(metric_contract.copy_mode),
),
metric_contract,
));
}
}
pub(super) async fn run_get_object_flow(
@@ -436,6 +498,7 @@ pub(super) async fn run_get_object_flow(
mod tests {
use super::get_object_strategy_range;
use super::*;
use futures_util::StreamExt;
use http::HeaderMap;
use rustfs_ecstore::store_api::ObjectOptions;
use rustfs_io_core::IoChunk;
@@ -483,47 +546,53 @@ mod tests {
assert_eq!(strategy_range.end, 511);
}
#[test]
fn classify_get_object_midstream_error_maps_expected_variants() {
assert_eq!(
classify_get_object_midstream_error(&std::io::Error::new(std::io::ErrorKind::UnexpectedEof, "eof")),
"unexpected_eof"
);
assert_eq!(
classify_get_object_midstream_error(&std::io::Error::new(std::io::ErrorKind::BrokenPipe, "closed")),
"channel_closed"
);
assert_eq!(
classify_get_object_midstream_error(&std::io::Error::other("bitrot verification failed")),
"bitrot"
);
assert_eq!(
classify_get_object_midstream_error(&std::io::Error::other("decode chunk failed")),
"decode"
);
}
#[tokio::test]
async fn instrument_get_object_chunk_stream_preserves_payload() {
let request_context = sample_request_context();
async fn materialize_chunk_stream_before_commit_buffers_small_payload_in_memory() {
let chunk_stream: BoxChunkStream = Box::pin(futures_util::stream::iter(vec![
Ok(IoChunk::Shared(bytes::Bytes::from_static(b"hello"))),
Ok(IoChunk::Shared(bytes::Bytes::from_static(b" world"))),
]));
let mut instrumented = instrument_get_object_chunk_stream(
&request_context,
chunk_stream,
GetObjectChunkPath::Direct,
rustfs_io_metrics::CopyMode::SharedBytes,
11,
);
let mut body = materialize_chunk_stream_before_commit_with_threshold(chunk_stream, 11, 8 * 1024, 1024)
.await
.unwrap()
.unwrap();
let mut collected = Vec::new();
while let Some(chunk) = instrumented.next().await {
collected.extend_from_slice(chunk.unwrap().as_bytes().as_ref());
while let Some(chunk) = body.next().await {
collected.extend_from_slice(&chunk.unwrap());
}
assert_eq!(collected, b"hello world");
}
#[tokio::test]
async fn materialize_chunk_stream_before_commit_falls_back_for_large_payload() {
let chunk_stream: BoxChunkStream = Box::pin(futures_util::stream::iter(vec![
Ok(IoChunk::Shared(bytes::Bytes::from_static(b"hello"))),
Ok(IoChunk::Shared(bytes::Bytes::from_static(b" world"))),
]));
// When payload exceeds the in-memory threshold an error is returned so the
// caller can fall back to the legacy reader path rather than spooling to disk.
let err = materialize_chunk_stream_before_commit_with_threshold(chunk_stream, 11, 8 * 1024, 4)
.await
.unwrap_err();
assert_eq!(err.streamed_bytes, 0);
assert_eq!(err.source.kind(), std::io::ErrorKind::Other);
}
#[tokio::test]
async fn materialize_chunk_stream_before_commit_rejects_short_body() {
let chunk_stream: BoxChunkStream =
Box::pin(futures_util::stream::iter(vec![Ok(IoChunk::Shared(bytes::Bytes::from_static(b"hello")))]));
let err = materialize_chunk_stream_before_commit_with_threshold(chunk_stream, 11, 8 * 1024, 1024)
.await
.unwrap_err();
assert_eq!(err.streamed_bytes, 5);
assert_eq!(err.source.kind(), std::io::ErrorKind::UnexpectedEof);
}
}