Commit Graph

50 Commits

Author SHA1 Message Date
唐小鸭 3792fed827 fix(replication): madmin reset/diff wire compat and config validation (#5799)
* fix(admin): align replication-reset responses with madmin ResyncTargetsInfo shape

The replication-reset and replication-reset-status responses serialized
their shell as "Targets" and per-target fields in PascalCase, while
madmin-go ResyncTargetsInfo/ResyncTarget expect the "target" shell key
and lowercase field tags (arn/resetid/resyncStatus/replicationCount/
completedReplicationSize/failedReplicationCount/failedReplicationSize).
Go json decoding is case-insensitive per field, but Targets vs target,
Status vs resyncStatus and the size/count key names cannot match, so
mc replicate resync decoded empty results.

Rename the serde tags to the exact madmin wire shape, keep the
ResetBeforeDate/Error RustFS extension keys (unknown keys are ignored
by Go decoders), pin the shape with a snapshot unit test, and update
the e2e client DTO to decode the madmin shape.

* fix(admin): stream bare madmin DiffInfo documents from replication diff

POST /v3/replication/diff returned a single enveloped object
({Entries, IsTruncated, ScannedVersions}) while madmin-go
BucketReplicationDiff decodes the body with a json.Decoder loop over
bare DiffInfo documents. The envelope decoded as exactly one DiffInfo
with an empty object, so mc replicate diff printed a phantom empty row
instead of the real backlog.

Emit one DiffInfo JSON document per line by default, using the exact
madmin json tags (object/versionId/rStatus/deletemarker/lastModified;
Size stays as a RustFS extension key that Go decoders ignore). The
enveloped shape moves to the opt-in ?aggregate=true RustFS extension,
which remains the only carrier of scan-coverage metadata; a truncated
default-mode scan is surfaced via a warn tracing event instead of
in-stream. Pin both shapes with unit tests and tighten the e2e helper
to reject any envelope in the stream.

* feat(replication): validate replication config structure before persisting

PutBucketReplication accepted structurally invalid configurations that
MinIO's replication.Config.Validate rejects: empty or oversized rule
lists, duplicate or negative rule priorities, over-long rule IDs,
filters carrying more than one of Prefix/Tag/And, and delete marker
replication enabled on tag-filtered rules. Such configs persisted
silently and later produced undefined routing (e.g. ambiguous priority
ties) instead of failing the PUT.

Add validate_replication_config_structure as a pure function in
rustfs-replication (limits documented as constants), surface it through
the ecstore api facade, and run it first in the PUT capability gate so
defects are named before any metadata write. Missing Priority counts as
zero for the uniqueness check, matching Go's zero-value semantics. The
self-target rejection deliberately stays at set-remote-target, where the
endpoint is known; a config can never reference a self-pointing ARN.
Document the rule-level Destination.StorageClass contract (use the
remote target's storage_class instead) and renumber the acceptance
matrix e2e to unique priorities, which MinIO would also require.

* test(replication): pin duplicated wire types with boundary reconciliation tests

rustfs-filemeta (xl.meta disk format) and rustfs-replication (MRF/resync
persistence format) deliberately each own ReplicationStatusType,
VersionPurgeStatusType and ReplicationState; the boundary converts
between them via as_str(), whose From<&str> impls fall back to Empty on
unknown tokens — a variant added on one side silently degrades to Empty
on the other.

Add reconciliation tests in replication_filemeta_boundary: exhaustive
matches with no wildcard arm on both sides of both enums (a new variant
fails compilation until the mapping is reconsidered), string-token
round-trip asserts (a token the other side does not recognize fails
instead of quietly becoming Empty), and a full-field ReplicationState
round-trip. Cross-reference the tests from both type definitions.
Struct drift was already compile-guarded by the exhaustive struct
literals in the conversion functions.

* docs(replication): define split completion criteria and milestone sequence

The ecstore replication split plan had no completion measure — the
boundary scaffolding risked ossifying because nothing said when the
migration counts as done. Record the criteria in the module inventory:
done means the Required Contracts table's 'Current dependency to
remove' column is empty; the end state moves pool/resyncer/state into
crates/replication, with the boundary micro-files dissolving as code
crosses the crate line (batch-merging them beforehand is explicitly
rejected — the guard scripts anchor on their file names, so merging is
churn with zero functional gain; only datatypes.rs can retire early).

Sequence the remaining work as M2 (resyncer pure decision logic, after
the oversized function splits) → M3 (worker runtime, highest risk,
last) → M4 (retire boundaries and guard entries). Refresh the stale
first-step text — the event sink / runtime contracts already landed —
and update the split-plan status table accordingly.

* fix(replication): align structural validator with MinIO semantics after adversarial review

Three interop corrections found by adversarial review of the new
structural validator, plus review fallout fixes:

- Delete-marker replication is now rejected only for a direct Filter.Tag,
  not for tags inside Filter.And — MinIO's validator only inspects the
  direct tag, and mc replicate add --tags "k1=v1&k2=v2" (delete-marker
  replication on by default) puts multiple tags into And.Tags, so the
  stricter check rejected mc-generated configs MinIO accepts.
- Rule ID length is measured in bytes (Go len semantics), not chars —
  a 255-char multibyte ID must not round-trip into a config MinIO
  rejects.
- An empty <Tag/> element (no key) counts as absent, matching MinIO's
  Tag.IsEmpty(); console form serializers emit empty tags, which would
  otherwise trip the exactly-one-of and delete-marker checks.

Also: repair the store-uninitialized PUT test whose empty-rules fixture
now (correctly) fails structural validation before reaching the store
lookup; pin the previously untested startTime madmin key in the
reset-status shape test; and signal a truncated default-mode diff scan
via the x-rustfs-replication-diff-truncated response header — the bare
madmin stream has no envelope, so a truncated scan was otherwise
indistinguishable from a complete healthy one (madmin/mc ignore unknown
headers).

* test(e2e): activate SSE-S3 replication contract and pin resync fail-closed path

The SSE-S3 replication contract e2e was ignored under backlog#1291
(silent plaintext replication); the fail-closed gate in
replication_target_boundary.rs closed that hole, so the ignore reason
expired. Un-ignore the test — it now pins the current fail-closed
contract (FAILED status, failure event, readable encrypted source,
stable absence of all target versions), verified green.

Add test_bucket_replication_sse_s3_resync_stays_fail_closed: drives the
existing-object resync path (PUT ?replication-reset) over a FAILED
SSE-S3 object and asserts the resync generation reaches a terminal
state without ever materializing a target version, with the
stays-absent window also spanning fast-scanner heal cycles. The new
start_bucket_replication_reset helper doubles as the madmin
ResyncTargetsInfo shape assertion (target[0].arn/resetid) for the
reset-start response.

Refresh the stale nextest count commentary (the module is at 20 fast +
36 nightly = 56 tests by cargo nextest list; the SSE-S3-ignored note no
longer holds).
2026-08-07 22:30:12 +08:00
cxymds 380ec74ece fix(replication): persist force-delete handoff state (#5641)
* fix(replication): persist force-delete handoff state

* fix(arch): route force-delete config access through boundary

* style: format force-delete imports

---------

Co-authored-by: Zhengchao An <anzhengchao@gmail.com>
2026-08-03 01:44:28 +00:00
houseme 035ce5d784 feat(obs): add bounded metrics dimensions (#5645)
* feat(obs): add drive topology detail metrics

Expose additive drive info, topology, state, and per-drive API metrics while preserving the existing drive metric label sets.

Backlog: rustfs/backlog#1655

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): preserve suspect drive runtime state

Keep suspect as a bounded drive runtime state and avoid all-zero runtime_state samples for that storage health state.

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): skip unknown drive inode samples

Avoid exporting zero inode gauges for missing or stale drive snapshots and ignore zero-count API latency buckets.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add scanner source work detail metrics

Expose additive scanner source and cycle work metrics with bounded server/source/state labels while leaving the existing aggregate scanner metrics unchanged.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add ilm action detail metrics

Expose additive ILM action/state task metrics with a server label while preserving the existing aggregate ILM series.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add delivery target server metrics

Expose additive audit and notification delivery target metrics with server labels and extend removed-target tombstones for the server-aware series.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add replication target flow metrics

Expose additive bucket replication target sent and failed-flow metrics while preserving existing bucket aggregates and target backlog series.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add request server metrics

Expose additive API request metrics with server labels while preserving the existing request and traffic metric label sets.

Co-Authored-By: heihutu <heihutu@gmail.com>

* style(obs): apply rustfmt to metrics changes

Apply rustfmt output to the metrics dimension changes without altering behavior.

Co-Authored-By: heihutu <heihutu@gmail.com>

* style(obs): reuse audit target label constant

Use the exported audit target_id label constant for legacy audit target metrics.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): populate drive disk metrics

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add scanner bucket drive result metrics

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(obs): add replication proxy server metrics

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): address metric liveness review

Use checked division for drive API latency aggregation and keep recovered drive, scanner current-cycle, replication flow, audit target, and notification target series from retaining stale values.

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): address metric dimension review

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): address additional metric review

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): count drive calls at start

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): address metrics dimension review

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): address dimension review gaps

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): address scanner review follow-ups

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): address runtime review follow-ups

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): reduce disk metric contention

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): address runtime review follow-ups

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(metrics): retire stale dimension series

Co-Authored-By: heihutu <heihutu@gmail.com>

---------

Co-authored-by: heihutu <heihutu@gmail.com>
2026-08-03 09:03:34 +08:00
cxymds 114b2420a2 feat(admin): expose versioned replication capabilities (#5631)
* feat(admin): expose replication capabilities

* fix(admin): route replication capabilities through facades
2026-08-02 22:54:38 +08:00
cxymds 2cc7443067 feat(replication): bound DeleteObjects queue admission (#5637)
feat(replication): batch DeleteObjects queue admission
2026-08-02 21:06:54 +08:00
cxymds 921ddef2c7 fix(lifecycle): bind delete replication admission (#5621) 2026-08-02 19:27:39 +08:00
cxymds c1955a8498 fix(replication): harden live delete admission (#5599) 2026-08-02 12:52:11 +08:00
houseme fbec33bd29 Expose target-scoped durable MRF backlog metrics (#5584)
* feat(replication): expose target durable mrf backlog

Add target ARN attribution to durable MRF entries and surface target-scoped durable backlog metrics without changing existing bucket-only metric labels.

Keep legacy MRF files bucket-only by defaulting missing targetARNs to an empty list, and expose target snapshots through an additive API so existing DurableMrfBacklogSummary callers remain source-compatible.

Co-Authored-By: heihutu <heihutu@gmail.com>

* feat(replication): expose runtime target backlog (#5586)

Track runtime replication backlog by target ARN for regular, large, delete, and MRF admission paths while preserving the existing bucket-level backlog semantics.

Add target-scoped current backlog metrics and merge them with durable target backlog snapshots for observability.

Co-authored-by: heihutu <heihutu@gmail.com>

---------

Co-authored-by: heihutu <heihutu@gmail.com>
2026-08-01 17:31:50 +00:00
唐小鸭 c8016cbcdb fix(replication): enforce bucket replication switches (#5449)
* fix(replication): enforce bucket replication switches

* fix(replication): satisfy delete admission clippy lint

* fix(replication): restore MinIO tag filter behavior

---------

Co-authored-by: Zhengchao An <anzhengchao@gmail.com>
Co-authored-by: cxymds <cxymds@gmail.com>
2026-08-01 15:03:57 +00:00
cxymds 7320d7fab2 fix(replication): make resync starts atomic (#5215) 2026-07-25 08:58:06 +08:00
cxymds 45b675c641 fix(replication): report authoritative backlog metrics (#5209) 2026-07-25 01:32:50 +08:00
Zhengchao An 13f8768e7f feat(ecstore): make replication timing intervals env-overridable for tests (#4680)
Introduce RUSTFS_REPL_HEALTH_CHECK_INTERVAL_MS, RUSTFS_REPL_MRF_FLUSH_INTERVAL_MS
and RUSTFS_REPL_RESYNC_POLL_MAX_MS so tests and operators can shorten the
replication background loops. Defaults are unchanged; invalid values fall
back with a warn and values below 10ms are clamped to avoid busy-spin.
Add replication_fast_env() e2e helper (backlog#1147 repl-4).
2026-07-10 21:57:43 +08:00
Zhengchao An 3d4fb532e5 refactor(replication): own filemeta wire contracts (#4254) 2026-07-04 06:39:54 +08:00
Zhengchao An 8a0617865b refactor(replication): route app storage through ecstore (#4251) 2026-07-04 02:00:28 +08:00
Zhengchao An 4883d3eb50 refactor(replication): route runtime facades through ecstore (#4248) 2026-07-03 23:49:24 +08:00
Zhengchao An 7cc470cb08 refactor(replication): isolate ecstore boundary imports (#4247) 2026-07-03 23:18:27 +08:00
Zhengchao An 769549dd2c refactor(replication): isolate status contract boundaries (#4236) 2026-07-03 19:40:49 +08:00
Zhengchao An 7001e53546 refactor(replication): isolate stats and app contract boundaries (#4235)
* refactor(replication): isolate stats boundary adapters

* refactor(replication): route app contracts through storage boundary
2026-07-03 18:48:26 +08:00
Zhengchao An aecac5c0ae refactor(replication): isolate object decision contracts (#4219) 2026-07-03 11:36:14 +08:00
Zhengchao An c260a2f20f refactor(replication): isolate queue boundary adapters (#4216) 2026-07-03 09:43:31 +08:00
Zhengchao An 48b70f6e4f refactor(replication): isolate resync boundary adapters (#4215) 2026-07-03 08:59:42 +08:00
Zhengchao An f0adabbe90 refactor(replication): move object config bridge (#4188)
* refactor(replication): move object config bridge

* refactor(replication): centralize log constants
2026-07-02 16:32:02 +08:00
Zhengchao An 473132f36b refactor(replication): move stats contracts into crate (#4168) 2026-07-02 12:14:36 +08:00
Zhengchao An 8ce0f824cf refactor(replication): move queue contracts into crate (#4167) 2026-07-02 11:40:49 +08:00
Zhengchao An 2cf856a51d Move replication delete worker contract into crate (#4166)
refactor(replication): move delete worker contract into crate
2026-07-02 11:03:24 +08:00
Zhengchao An df6ce41acd Move replication operation contracts into crate (#4165)
refactor(replication): move operation contracts into crate
2026-07-02 10:48:02 +08:00
Zhengchao An b57820a486 refactor(replication): move config contracts into crate (#4162) 2026-07-02 10:17:48 +08:00
Zhengchao An e80391f1aa refactor(replication): extract resync contracts crate (#4154) 2026-07-02 01:30:01 +08:00
Zhengchao An 22cbd41ee9 refactor(ecstore): add replication owner bridges (#4141)
* refactor(ecstore): add replication owner bridges

* fix(ecstore): tighten replication owner guards
2026-07-01 21:51:26 +08:00
Zhengchao An c0dfc260ca refactor(ecstore): add replication object bridge (#4140) 2026-07-01 19:24:20 +08:00
Zhengchao An e5ab6ee8e4 refactor(ecstore): make replication facade explicit (#4139) 2026-07-01 18:54:05 +08:00
Zhengchao An fc806679fb refactor(ecstore): add replication work bridges (#4137) 2026-07-01 18:37:01 +08:00
Zhengchao An 722c0805ce refactor(ecstore): route replication errors (#4127) 2026-07-01 08:13:57 +08:00
Zhengchao An e917f1f4be refactor(ecstore): route replication filemeta contracts (#4122) 2026-07-01 01:01:01 +08:00
Zhengchao An da7f421f69 refactor(ecstore): route replication storage contracts (#4118) 2026-06-30 20:12:06 +08:00
Zhengchao An 20919f0b97 refactor(ecstore): route replication bandwidth wrapping (#4113) 2026-06-30 15:42:04 +08:00
Zhengchao An fb7bcfb541 refactor(replication): add tagging boundary (#4081) 2026-06-30 03:39:17 +08:00
Zhengchao An d3a84395dc refactor(replication): add msgp boundary (#4076) 2026-06-30 02:09:14 +08:00
Zhengchao An c17df6e3d3 refactor(replication): add lock boundary (#4075) 2026-06-30 00:18:08 +08:00
Zhengchao An 50ca1c2cbe refactor(replication): add config store boundary (#4073) 2026-06-30 00:03:45 +08:00
Zhengchao An f8c70e017d refactor(replication): add versioning boundary (#4071) 2026-06-29 23:52:36 +08:00
Zhengchao An 04e8c1bb6b refactor(replication): add metadata boundary (#4070) 2026-06-29 23:24:39 +08:00
Zhengchao An f1ec0bd765 refactor(replication): add target boundary (#4069) 2026-06-29 22:59:10 +08:00
Zhengchao An c94f2981d7 refactor(replication): add event sink boundary (#4068) 2026-06-29 22:46:06 +08:00
Zhengchao An 13474ec1de refactor(replication): add runtime boundary (#4067) 2026-06-29 22:34:10 +08:00
安正超 a0e7957e92 refactor: prune admin app compat re-exports (#3697) 2026-06-21 20:28:43 +08:00
LeonWang0735 7f1cdaedad feat(replication): add bandwidth-aware reporting for bucket replication metrics (#2141) 2026-03-15 09:03:10 +08:00
weisd cd1e244c68 Refactor: Introduce content checksums and improve multipart/object metadata handling (#671)
* feat:  adapt to s3s typed etag support

* refactor: move replication struct to rustfs_filemeta, fix filemeta transition bug

* add head_object checksum, filter object metadata output

* fix multipart checksum

* fix multipart checksum

* add content md5,sha256 check

* fix test

* fix cargo

---------

Co-authored-by: overtrue <anzhengchao@gmail.com>
2025-10-20 23:46:13 +08:00
weisd 90f21a9102 refactor: Reimplement bucket replication system with enhanced architecture (#590)
* feat:refactor replication

* use aws sdk for replication client

* refactor/replication

* merge main

* fix lifecycle test
2025-09-26 14:27:53 +08:00
houseme 5826396cd0 refactor: Restructure project layout and clean up dependencies (#30)
This commit introduces a significant reorganization of the project structure to improve maintainability and clarity.

Key changes include:
- Adjusted the directory layout for a more logical module organization.
- Removed unused crate dependencies, reducing the overall project size and potentially speeding up build times.
- Updated import paths and configuration files to reflect the structural changes.
2025-07-02 19:33:12 +08:00