Flaky test dashboard #20437

This issue was generated automatically by Claude Code (Anthropic's AI coding agent) running a scheduled CI-triage routine on behalf of @FrankChen021. Analysis and suggested fixes are AI-produced; please verify before acting on them.

This dashboard tracks failed push-triggered jobs on master (Static Checks CI, CodeQL, Unit & Integration tests CI). Each problem has its own Flaky test tracking issue with the root cause, suggested fix and full occurrence history. The routine rewrites this body every day and never comments here.

Last updated: 2026-10-09 (covers master commits through 2026-10-08)

Open problems

Sorted by failures in the last 14 days, then by total failures.

Yesterday (2026-10-08)

• 3 of 16 master commits had failed jobs: bd61838 (build: add the Maven wrapper and use it in CI #20496, 20 jobs), 51b66c7 (build(deps): bump com.google.api:gax from 2.85.0 to 2.86.0 #20516, 1 job) and bbeb731 (build(deps): bump aws.sdk.v2.version from 2.55.1 to 2.55.4 #20509, 1 job).
• New: none.
• Recurred: CI infra: Maven cannot resolve parent POM org.apache:apache (${repoOrgUrl}) #20436 (reopened: org.apache:apache:pom:40 not found in Central on all 3 attempts of the pre-resolve step added by ci: resolve Apache parent POM before worker builds #20459), Flaky test: KinesisFaultToleranceTest.test_supervisorRecovers_afterSuspendResume_withEmptyShards #20433 (COUNT(*) 2001 vs expected 3000), Regression: ProtobufEventSerializer test compile break after #20495 (ProbufData rename) #20517 (bd61838 merged before fix: Fix broken compile #20506 and inherited the ProbufData compile break; not a recurrence, issue stays closed).
• Failed once and passed on retry, so the job did not fail and they are not tracked: ConcurrentGrouperTest.testGrouperTimeout (2 commits: 971f21b, eb43fd0; DirectoryNotEmpty on the JUnit temp dir), EmbeddedDimensionValueSetShardSpecTest.test_pruning_verifiedBySegmentScanMetric (723d3a2) and test_twoSupervisors_withPartitionDimensionValues_segmentsPublishAndQueriesSucceed (e9cc7e8; both timed out waiting for an event after 10 s), KinesisIndexTaskTest.testRestore (51b66c7; 120 s timeout), TaskQueueTest.testGetTaskStatus_successfulTask (51b66c7; RUNNING instead of SUCCESS), KafkaSupervisorTest.testCheckpointForInactiveTaskGroup (971f21b; see the older human-filed Transient failure of KafkaSupervisorTest.testCheckpointForInactiveTaskGroup #6296), ReservoirSegmentSamplerTest.testSegmentsFromAllServersAreEquallyLikelyToBePicked (723d3a2; 572 vs expected 625), EmbeddedKafkaSupervisorTest.test_runKafkaSupervisorWithHeaderFiltering (1319898, before test: fix flaky EmbeddedKafkaSupervisorTest header filtering test #20519 merged).
• No regressions caused by a commit. bd61838's 20 failed jobs come from the test: remove protobuf-dynamic dependency #20495 compile break tracked in Regression: ProtobufEventSerializer test compile break after #20495 (ProbufData rename) #20517, which fix: Fix broken compile #20506 fixed later the same day.

Recently resolved (last 14 days)