Commit Graph
6130 Commits
Author SHA1 Message Date
1b3744a1da test(perf): align GET attribution harness with backlog 2093 (#6912)
Co-authored-by: heihutu <[email protected]>
2026-08-30 20:28:28 +00:00
9244eb36ed test(e2e): route cluster volume endpoints through fault proxy (#6909)
Co-authored-by: heihutu <[email protected]>
2026-08-30 20:17:16 +00:00
442298d5f7 test(ecstore): prove in-flight prefetch cancellation (#6904)
Co-authored-by: heihutu <[email protected]>
2026-08-30 19:17:44 +00:00
be7d35d441 perf(get): release disk permits for buffered bodies (#6906)
Co-authored-by: heihutu <[email protected]>
2026-08-30 19:10:26 +00:00
唐小鸭andGitHub ec1cd606d3 fix(replication): surface object-lock denied purges and back off heal retries (#6900) 2026-08-30 18:59:55 +00:00
16af688a7a fix(rpc): reject unsigned v2 control mutations (#6905)
Co-authored-by: heihutu <[email protected]>
2026-08-30 18:17:43 +00:00
唐小鸭andGitHub 37b23a16da fix(replication): verify replica integrity and default to plain signed payloads (#6895) 2026-08-31 01:43:45 +08:00
006e9b7d28 test(e2e): cover four-node four-drive cluster topology (#6902)
Co-authored-by: heihutu <[email protected]>
2026-08-30 17:32:36 +00:00
d214c27583 perf(ecstore): consolidate non-inline read planning (#6892)
Co-authored-by: heihutu <[email protected]>
2026-08-30 17:15:48 +00:00
GatewayJandGitHub 8fd364a99c feat(s3-tables): support object-backed table rename (#6899) 2026-08-31 00:30:49 +08:00
c2d8488728 docs(architecture): reconcile generation contract (#6901)
Co-authored-by: heihutu <[email protected]>
2026-08-31 00:20:39 +08:00
唐小鸭andGitHub 1370434f3a fix(scanner): unblock quota usage baseline on never-converged sites (#6896) 2026-08-31 00:20:04 +08:00
唐小鸭andGitHub 5dde2c188c fix(replication): retry failed multipart aborts on bounded backoff (#6897) 2026-08-31 00:19:49 +08:00
2f9c75d04f perf(ecstore): reuse prepared metadata across pools (#6889)
Co-authored-by: heihutu <[email protected]>
2026-08-30 16:15:10 +00:00
唐小鸭andGitHub 9ee7b1221d fix(admin): replicate user secret-key rotation to peer sites (#6893) 2026-08-30 23:32:07 +08:00
Zhengchao AnandGitHub fcc3c7fb6b test(s3): promote passing compatibility cases (#6891) 2026-08-30 21:56:17 +08:00
Zhengchao AnandGitHub 01dc55ee5b docs(security): add unsigned presign header lesson (#6894) 2026-08-30 21:55:47 +08:00
3d24526704 fix(ecstore): preserve parity reserves for data-only GET (#6888)
fix(ecstore): hedge data-only GET with parity

Route the opt-in data-shards-only lockstep path through the bounded parity race and preserve deferred parity reserves across canceled hedges.

Co-authored-by: heihutu <[email protected]>
2026-08-30 20:16:32 +08:00
51532e19fb test(ecstore): cover multipart snapshot overwrite race (#6887)
test(ecstore): cover multipart GET overwrite snapshot

Co-authored-by: heihutu <[email protected]>
2026-08-30 12:15:02 +00:00
931ff60182 test(ci): refresh cluster nightly selection (#6886)
Co-authored-by: heihutu <[email protected]>
2026-08-30 12:10:07 +00:00
07212c4e26 perf(ecstore): gate quorum-aware GET early stop (#6885)
* perf(ecstore): add gated two-phase GET metadata reads

Co-Authored-By: heihutu <[email protected]>

* fix(ecstore): require data-shard coverage for read plans

Co-Authored-By: heihutu <[email protected]>

* perf(ecstore): avoid inline overhead in read plan rollout

Co-Authored-By: heihutu <[email protected]>

* perf(ecstore): accept quorum-complete read candidates

Co-Authored-By: heihutu <[email protected]>

---------

Co-authored-by: heihutu <[email protected]>
2026-08-30 09:33:35 +00:00
GatewayJandGitHub 4932af080b feat(s3select): expand typed JSON source paths (#6864) 2026-08-30 06:46:36 +00:00
GatewayJandGitHub d6f9a7c462 feat(table-catalog): vend credentials from LoadTable (#6878)
* feat(table-catalog): vend credentials from LoadTable

* fix(table-catalog): preserve entry-relative metadata paths
2026-08-30 06:33:28 +00:00
7345b49cf6 perf(ecstore): gate GET metadata timing when metrics off (#6879)
Co-authored-by: heihutu <[email protected]>
2026-08-30 05:43:08 +00:00
housemeandGitHub 4753e35035 chore(deps): update flake.lock (#6880) 2026-08-30 13:14:13 +08:00
GatewayJandGitHub 96239fc034 feat(s3select): report uncompressed input byte metrics (#6865) 2026-08-30 04:07:20 +00:00
hectorandGitHub b428875bed fix(ci): stabilize tier MQTT bootstrap on shared runner (#6877)
* fix(ci): isolate s3 compat temp file paths

* fix(ci): use rooted auto-testing s3 temp fix

* fix(ci): follow auto-testing main after temp-path merge

* fix(ci): stabilize tier mqtt bootstrap on shared runner
2026-08-30 11:14:28 +08:00
Zhengchao AnandGitHub cf362282f0 fix(test): serialize transition matrix tests under nextest (#6874)
The transition_matrix_tests use #[serial_test::serial] which has no
effect under nextest (each test runs in a separate process). When running
alongside thousands of other ecstore tests, the shared metadata cache
generation counter can race, causing intermittent 'metadata read should
publish the generation under test' panics.

Add both tests to the ecstore-serial-flaky test group in both default
and ci nextest profiles so they run single-threaded.
2026-08-30 10:42:23 +08:00
Zhengchao AnandGitHub b2a2e637a5 fix(ci): refresh Linux full E2E selection (#6875) 2026-08-30 10:42:14 +08:00
cxymdsandGitHub 0c18012442 fix(admin): version remote target credential capabilities (#6876) 2026-08-30 10:42:10 +08:00
ee39e4fccb fix(scanner): own publication mutations through storage drain (#6867)
* fix(scanner): own publication mutations through storage drain

Co-Authored-By: heihutu <[email protected]>

* fix(storage): remove unused rename data shim

Co-Authored-By: heihutu <[email protected]>

---------

Co-authored-by: heihutu <[email protected]>
2026-08-30 02:39:07 +00:00
90ab2e24c3 perf(ecstore): reuse local fd metadata snapshots (#6868)
* perf(ecstore): reuse local fd metadata snapshots

Cache the validated shard length beside each reusable descriptor so read hits avoid a repeated fstat while retaining generation and mutation invalidation semantics.

Co-Authored-By: heihutu <[email protected]>

* fix(ecstore): pass cached entry to fd cache

Co-Authored-By: heihutu <[email protected]>

---------

Co-authored-by: heihutu <[email protected]>
2026-08-30 09:08:44 +08:00
cxymdsandGitHub 21e5b3dc64 fix(ecstore): require durable decommission ledger format (#6871) 2026-08-30 08:47:09 +08:00
cxymdsandGitHub 1e8c8d4cd5 feat(replication): support temporary target credentials (#6860) 2026-08-30 08:44:34 +08:00
ff3ad30f0c fix(scanner): bound publication proof retries on main (#6870)
* fix(scanner): retain completed publication candidates

* fix(scanner): export publication activity helper

* test(ecstore): retain activity snapshot across retries

* fix(scanner): rebase publication proof retry onto main

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): resolve publication proof retry conflicts

Co-Authored-By: heihutu <[email protected]>

---------

Co-authored-by: Henry Guo <[email protected]>
Co-authored-by: heihutu <[email protected]>
2026-08-29 22:41:36 +00:00
47a3f5ef01 perf(ecstore): converge disk metric atomic loads (#6866)
Use the seqlock version as the publication fence and keep payload reads relaxed while validating the final version. This reduces ordering overhead in disk metric recording and snapshot collection without changing the rolling-window contract.

Co-authored-by: heihutu <[email protected]>
2026-08-29 20:38:59 +00:00
a22fa7461d perf(put): adapt eager threshold to concurrency (#6863)
Co-authored-by: heihutu <[email protected]>
2026-08-29 20:21:05 +00:00
814ab5bbf3 fix(ecstore): classify system metadata failures (#6862)
fix(ecstore): classify system metadata volume failures

Preserve retryable quorum errors when system metadata reads or writes encounter missing volumes, and cover the create-bucket data-usage path with regressions.

Co-authored-by: heihutu <[email protected]>
2026-08-29 19:48:50 +00:00
498205b7ec fix(ecstore): keep 1MiB GET off mid-size reader (#6861)
Co-authored-by: heihutu <[email protected]>
2026-08-29 19:39:51 +00:00
c235f7c05d fix(scanner): retain usage across transient peer failures (#6859)
* test(scanner): cover bucket drive guard lifecycle

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): recover usage floor from fenced backups

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): retry transient activity probes

Retry one failed scanner activity probe after a bounded reconnect when the failure is transport-like or timed out. Keep protocol and response validation failures fail-closed.

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): retain post-scan observations

Preserve a complete scanner walk as a non-converged observation when the final activity probe is unavailable. Advance the cycle as partial without acknowledging dirty usage.\n\nCo-Authored-By: heihutu <[email protected]>

* fix(scanner): classify publication lease deferrals

Distinguish persistence budget and lease deadline deferrals from unavailable activity baselines, and ensure lease-gate deferrals update usage metrics. Keep the fixed lease gate fail-closed while storage-owned commit scope work remains pending.

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): recover usage floor from fenced backups

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): preserve publication lease defer reasons

Keep lease expiry and release failures distinct from activity baseline failures so scanner freshness metrics and cycle outcomes identify the publication barrier that blocked progress. Preserve fail-closed behavior.

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): reuse recovered usage baseline for publication

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): fence legacy usage floor fallback

Co-Authored-By: heihutu <[email protected]>

* fix(scanner): use typed activity timeout error

---------

Co-authored-by: heihutu <[email protected]>
2026-08-30 02:54:20 +08:00
cxymdsandGitHub 64cca79fbb feat(admin): expose remote target credential capability state (#6857) 2026-08-30 00:42:14 +08:00
759e1041bd feat(nix): add NixOS service module and client package (#6856)
Co-authored-by: heihutu <[email protected]>
2026-08-29 16:14:44 +00:00
GatewayJandGitHub 8055aeb1d4 test(s3select): cover SelectRequest XML root alias (#6820) 2026-08-29 14:14:12 +00:00
GatewayJandGitHub 79bd6fa862 fix(s3select): return encryption response headers (#6819) 2026-08-29 19:46:34 +08:00
fa0be5d271 fix(ci): split workflows and add perf version reporting (#6848)
* ci: pass package selector to test run steps (fix rc.3 fallback)

* fix(ci): split workflows and add perf version reporting

* fix(ci): enforce strict shared workflow order

---------

Signed-off-by: houseme <[email protected]>
Co-authored-by: houseme <[email protected]>
2026-08-29 19:32:19 +08:00
78cb142c91 fix(s3): accept empty put without content length (#6849)
Co-authored-by: heihutu <[email protected]>
2026-08-29 19:31:33 +08:00
hectorandGitHub 5fa3d2a682 ci: pass package selector to test run steps (fix rc.3 fallback) (#6846) 2026-08-29 17:34:04 +08:00
hectorandGitHub fd8ddf0a02 fix: remove redundant --repo flag in preview release cleanup (#6847)
The check_preview_release_workflow.sh script uses exact line matching
(grep -Fxq) to verify the cleanup-preview-releases job contains:

  gh release delete "$preview_tag" --yes

The extra --repo flag is unnecessary in GitHub Actions context since
gh auto-detects the repository from GITHUB_REPOSITORY, and it causes
the Workflow Pin Report check to fail on all PRs.
2026-08-29 17:33:34 +08:00
Zhengchao AnandGitHub e1ea99ff06 fix(s3): return BadDigest for Content-MD5 mismatch (#6842) 2026-08-29 09:00:18 +00:00
唐小鸭andGitHub 11c6ee42ea fix(kms): restore persisted configuration after restart (#6821)
* fix(kms): restore persisted configuration after restart

* docs(kms): cover the reload route and startup load states

The admin contract matrix pins every dynamic KMS route for the rc and
console handoff, so the new POST /kms/reload needs a row there, and the
reload response reuses the configure snapshot shape rather than adding a
wire type. The observability runbook gains the operator procedure the
reload exists for: telling a load_failed startup apart from a server
that was never configured, and recovering without resubmitting secrets.
2026-08-29 16:21:22 +08:00