mirror of
https://github.com/rustfs/rustfs.git
synced 2026-07-26 08:18:18 +00:00
7c701d9f2c
Two ecstore test groups pass in isolation but flake under the full parallel nextest suite, producing CI noise on every in-flight PR (backlog #937): - store::bucket::tests::bucket_delete_* race make_bucket into InsufficientWriteQuorum because they share global disk-registry/lock-client state with other concurrently running tests. - bucket_lifecycle_ops::tests::concurrent_resend_same_part_commits_one_generation exceeds its already-maxed 60s lock-acquire deadline only when the suite saturates disk I/O. serial_test's #[serial] does not serialize these across nextest's per-test process boundary. Add a nextest test-group (max-threads = 1) that matches both groups so they run serialized, removing the load-induced flake. No test or production code changes; long-term global-singleton isolation is tracked separately. Co-authored-by: heihutu <heihutu@gmail.com>