Relationships
#4953 Test flake: v3 replication 'an owned sidecar whose owner's pid names another process stops' fails with 'the Litestream sidecar did not start in time' under load
Opened by zack · 11/20/2026· Shipped 11/20/2026
On main at df967e324 plus an unrelated unit-test-only change (713ca82da, swamp-club#4952), integration/datastore_v3_litestream_replication_test.ts:888 failed in two consecutive verify-build runs (run ids 0243d4f5-0163-46b7-ba27-d97be14447d2 and a1109318-523e-406d-a7a8-5a021d782e00) with V3ReplicationUnavailableError: swamp cannot replicate this v3 datastore: the Litestream sidecar did not start in time. Nothing was changed. Stack: LitestreamReplicator.#findOrStart (litestream_replicator.ts:417) via acquire, acquireV3Replication, takeV3WriterHold, holdV3Writer, writeCommand (datastore_v3_litestream_helpers.ts:257), from the test at line 906 inside withStandInOwner (line 800). Run alone with the litestream 0.5.17 binary from LitestreamBinaryProvider it passes 3 of 3, so it fails only under the full sharded suite's load. This is a different failure mode from the earlier flakes in the same test (swamp-club#4785, #4786, #4841, all shipped): the sidecar start deadline from TEST_TIMINGS appears too tight on a loaded host. The test should wait on the sidecar's readiness rather than a fixed start deadline, or use a start deadline that does not depend on host load.
Shipped
Click a lifecycle step above to view its details.
Sign in to post a ripple.