Skip to content

Flaky test report: committed-code failures on 2026-06-10 #291

Description

@andrross

Summary

10 distinct tests failed against committed code (Timer or Post Merge Action builds on main) in the 24 hours ending 2026-06-10T10:00 UTC. None reproduced locally with the original seed, consistent with non-deterministic (timing/environment-dependent) flakiness.

Summary Table (sorted by total builds affected)

# Test Builds Affected First Seen Trend Reproduced?
1 RemoteStoreStatsIT 217 2024-03-27 Worsening — recent months show increased failure rate No — passed with seed
2 MixedClusterClientYamlTestSuiteIT 148 2024-03-25 Worsening — recent months show increased failure rate N/A — requires multi-version cluster
3 SharedClusterSnapshotRestoreIT 108 2024-08-31 Worsening — recent months show increased failure rate No — passed with seed
4 SecureReactorNetty4HttpServerTransportTests 67 2025-02-19 Improving — recent months show decreased failure rate No — passed with seed
5 IngestFromKafkaIT 54 2025-10-15 Worsening — recent months show increased failure rate No — passed with seed
6 AggregationProfilerIT 26 2025-08-07 Worsening — recent months show increased failure rate No — passed with seed
7 SegmentReplicationUsingRemoteStoreIT 12 2024-07-15 Worsening — recent months show increased failure rate No — passed with seed
8 IndexRecoveryIT 12 2024-06-05 Worsening — recent months show increased failure rate No — passed with seed
9 AdmissionControlSingleNodeTests 8 2025-04-14 Stable — consistent low-level flakiness No — passed with seed
10 IngestPipelineFromKafkaIT 3 2026-03-30 Stable — consistent low-level flakiness No — passed with seed

Detailed Findings

1. RemoteStoreStatsIT.testNonZeroPrimaryStatsOnNewlyCreatedIndexWithZeroDocs

Field Value
Build #79997
Seed 98627D977CB6673C:5595A84B3AA9B0FA
Error java.lang.AssertionError
First failure 2024-03-27
Total builds affected 217
Recent monthly failures (2026) 2026-01=7, 2026-02=2, 2026-03=9, 2026-04=8, 2026-05=14, 2026-06=12
Reproduced locally No — passed with seed
Trend Worsening — recent months show increased failure rate

2. MixedClusterClientYamlTestSuiteIT.test {p0=cluster.health/10_basic/cluster health with closed index}

Field Value
Build #79975
Seed 6AD408F9F4AB03EB:E28037235A576E13
Error java.lang.AssertionError: Failure at [cluster.health/10_basic:182]: expected [2xx] status code but api [cluster.health] returned [408 Request Timeout] [{"cluster_name":"v3.6.1","status":"red","timed_o
First failure 2024-03-25
Total builds affected 148
Recent monthly failures (2026) 2026-01=1, 2026-04=9, 2026-05=11, 2026-06=13
Reproduced locally N/A — requires multi-version cluster
Trend Worsening — recent months show increased failure rate

3. SharedClusterSnapshotRestoreIT.testSnapshotFileFailureDuringSnapshot

Field Value
Build #80058
Seed CA37A560C385DC0F:F545090A3AB8BA12
Error `java.lang.AssertionError:
Expected: <0L>
 but: was <1L>` |

| First failure | 2024-08-31 |
| Total builds affected | 108 |
| Recent monthly failures (2026) | 2026-01=3, 2026-03=6, 2026-04=10, 2026-05=9, 2026-06=5 |
| Reproduced locally | No — passed with seed |
| Trend | Worsening — recent months show increased failure rate |

4. SecureReactorNetty4HttpServerTransportTests.testBadRequest

Field Value
Build #79964
Seed E43DC6930946AF27:F50C8D304839DBDC
Error reactor.core.Exceptions$ReactiveException: reactor.netty.http.client.PrematureCloseException: Connection prematurely closed BEFORE response
First failure 2025-02-19
Total builds affected 67
Recent monthly failures (2026) 2026-06=2
Reproduced locally No — passed with seed
Trend Improving — recent months show decreased failure rate

5. IngestFromKafkaIT.testAllActiveOffsetBasedLag

Field Value
Build #80024
Seed C12A3978C955D4CA:62CBB61E56B3475
Error java.lang.AssertionError
First failure 2025-10-15
Total builds affected 54
Recent monthly failures (2026) 2026-03=8, 2026-04=13, 2026-05=25, 2026-06=6
Reproduced locally No — passed with seed
Trend Worsening — recent months show increased failure rate

6. AggregationProfilerIT.testTopHitsAggregationFetchProfiling {p0={"search.concurrent_segment_search.enabled":"true"}}

Field Value
Build #80029
Seed 2ACBF530BFD46BA8:51059F0E421632E3
Error org.opensearch.action.search.SearchPhaseExecutionException: all shards failed
First failure 2025-08-07
Total builds affected 26
Recent monthly failures (2026) 2026-03=2, 2026-04=1, 2026-05=5, 2026-06=4
Reproduced locally No — passed with seed
Trend Worsening — recent months show increased failure rate

7. SegmentReplicationUsingRemoteStoreIT.testPrimaryStopped_ReplicaPromoted

Field Value
Build #79924
Seed 3A9E9A3DB968E533:2B969B5AA61A120B
Error java.lang.AssertionError: Count is 2 hits but 3 was expected. Total shards: 1 Successful shards: 1 & 0 shard failures:
First failure 2024-07-15
Total builds affected 12
Recent monthly failures (2026) 2026-04=3, 2026-05=4, 2026-06=1
Reproduced locally No — passed with seed
Trend Worsening — recent months show increased failure rate

8. IndexRecoveryIT.testOngoingRecoveryAndClusterManagerFailOver

Field Value
Build #80020
Seed B82047EE4FA4447F:AE78B75A37F1BB9E
Error java.lang.AssertionError: timed out waiting for green state
First failure 2024-06-05
Total builds affected 12
Recent monthly failures (2026) 2026-05=4, 2026-06=3
Reproduced locally No — passed with seed
Trend Worsening — recent months show increased failure rate

9. AdmissionControlSingleNodeTests.testAdmissionControlRejectionDisabledMode

Field Value
Build #80024
Seed C12A3978C955D4CA:10A2707B0AD35D4
Error java.lang.AssertionError: expected:<3> but was:<6>
First failure 2025-04-14
Total builds affected 8
Recent monthly failures (2026) 2026-03=1, 2026-06=1
Reproduced locally No — passed with seed
Trend Stable — consistent low-level flakiness

10. IngestPipelineFromKafkaIT.testCombinedScriptTransformations

Field Value
Build #79989
Seed 78A1D1604AFDB051:7F2405C1BA8BBD63
Error org.awaitility.core.ConditionTimeoutException: Condition org.opensearch.plugin.kafka.KafkaUtils$$Lambda/0x000000002a6bb978 was not fulfilled within 1 minutes.
First failure 2026-03-30
Total builds affected 3
Recent monthly failures (2026) 2026-03=1, 2026-06=2
Reproduced locally No — passed with seed
Trend Stable — consistent low-level flakiness

Methodology

  1. Queried gradle-check-* indices on the OpenSearch metrics cluster for test_status=FAILED in the past 24h, filtered to committed-code builds (invoke_type=Timer, git_reference=main or invoke_type=Post Merge Action).
  2. Queried full historical failure data across all build types (including PR builds) with monthly date histogram and cardinality aggregation on build_number.
  3. Extracted test seeds from Jenkins test reports (the SeedInfo.seed([...]) line in stack traces).
  4. Attempted local reproduction: ./gradlew :<module>:<task> --tests "<class>.<method>" -Dtests.seed=<SEED>.
  5. None of the 9 locally-runnable tests reproduced. The MixedClusterClientYamlTestSuiteIT requires a multi-version backward-compatibility cluster and could not be run locally with a simple gradle command.

Observations

  • None of the seeds are deterministic reproducers. This is expected for timing-dependent flakes — RandomizedRunner seeds control Random streams but not thread scheduling, network timing, GC pauses, or JIT warmup.
  • CI runner migration (April 2026): The move from m5.8xlarge to m7a.8xlarge correlates with upticks in MixedClusterClientYamlTestSuiteIT, SegmentReplicationUsingRemoteStoreIT, and IndexRecoveryIT.
  • RemoteStoreStatsIT is the most chronic offender (217 builds since March 2024) with consistent ~5-15 builds/month. No sign of improvement.
  • IngestFromKafkaIT is worsening rapidly (2 builds in Oct 2025 → 25 in May 2026).
  • SecureReactorNetty4HttpServerTransportTests had a spike in Sep 2025 (33 builds) but has been mostly quiet since. The June 2026 occurrence may be a recurrence.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions