Flaky test report: committed-code failures on 2026-04-01
Summary
11 test failures were detected across committed-code builds (Timer and Post Merge Action) in the past 24 hours. These map to 10 distinct failing tests (capped at 10 for this report). One test (ConcurrentSeqNoVersioningIT) reproduced locally with the original seed.
Failing Tests
1. ClientYamlTestSuiteIT — string profiler via global ordinals
Build : 73664
Seed : 2F166AF4E6F1B4F1
Error : field [profile.shards.0.aggregations.0.debug.segments_with_single_valued_ords] is not greater than [0]
Reproduced locally : No
First seen : 2024-03-25
Total unique builds affected : 350
Pattern : Persistent long-running flake since March 2024. Major spike in Sep 2024 (89 builds). Steady at 4–24 builds/month since. Stable/chronic.
2. MixedClusterClientYamlTestSuiteIT — string profiler via global ordinals
3. AwarenessAllocationIT — testThreeZoneOneReplicaWithForceZoneValueAndLoadAwareness
Build : 73640
Seed : D2E4E2BABCE53B0C
Error : Expected: <120> but: was <118>
Reproduced locally : Inconclusive (Gradle infra error)
First seen : 2024-08-31
Total unique builds affected : 135
Pattern : Quiet until Apr 2025, then consistent 3–18 builds/month. Trending upward. Worsening.
4. RecoveryWhileUnderLoadIT — testRecoverWhileUnderLoadAllocateReplicasRelocatePrimariesTest
Build : 73633
Seed : E6978E901AAC624
Error : NoSuchElementException: No value present
Reproduced locally : Inconclusive (Gradle infra error)
First seen : 2024-05-03
Total unique builds affected : 117
Pattern : Sporadic early on, major spikes in Apr–Jul 2025 (21–38 builds/month). Resurgence in Feb–Mar 2026 (8–11 builds/month). Worsening.
5. FullRollingRestartIT — testFullRollingRestart_withNoRecoveryPayloadAndSource
Build : 73652
Seed : 6E21A05A37A8A2E3
Error : replica shards haven't caught up with primary expected:<18> but was:<15>
Reproduced locally : No
First seen : 2024-10-11
Total unique builds affected : 104
Pattern : Dormant until Jul 2025, then massive spike (47 builds). Quieted, then resurgence in Feb–Mar 2026 (17–13 builds/month). Worsening.
6. AzureBlobStoreRepositoryTests — testWriteRead
Build : 73707
Seed : None available
Error : RepositoryVerificationException: path not accessible on cluster-manager node (BlobAlreadyExists/CONFLICT)
Reproduced locally : Skipped (requires Docker for Azure fixture)
First seen : 2024-04-29
Total unique builds affected : 79
Pattern : Persistent low-level flake since Apr 2024. Consistent 1–9 builds/month. Slight upward trend recently. Stable/chronic.
7. ConcurrentSeqNoVersioningIT — testSeqNoCASLinearizability
Build : 73690
Seed : C23880DF2C94CBC3
Error : ClusterHealthResponse has timed out
Reproduced locally : Yes — failure reproduced deterministically with seed C23880DF2C94CBC3
First seen : 2024-10-03
Total unique builds affected : 66
Pattern : Consistent 1–10 builds/month since Oct 2024. Spike in Feb 2026 (10 builds). Stable/chronic.
8. RestoreShallowSnapshotV2IT — testRestoreShallowSnapshotIndexAfterSnapshot
Build : 73696
Seed : BD4077D675BFCA75
Error : timed out waiting for green state
Reproduced locally : No
First seen : 2024-09-03
Total unique builds affected : 50
Pattern : Consistent 1–8 builds/month since Sep 2024. Quiet Mar–May 2025, then resumed. Stable/chronic.
9. TelemetryMetricsEnabledSanityIT — testGauge
Build : 73653
Seed : FA09294D2AF3FDAA
Error : expected:<2.0> but was:<3.0>
Reproduced locally : No
First seen : 2024-07-15
Total unique builds affected : 30
Pattern : Low frequency until Mar 2026, then sudden spike to 17 builds in a single month. Worsening.
10. WarmIndexSegmentReplicationIT — testPrimaryStopped_ReplicaPromoted
Build : 73668
Seed : 8644DC5BA7FFD3B5
Error : RefreshFailedEngineException: Refresh failed (CorruptIndexException: misplaced codec footer)
Reproduced locally : Inconclusive (Gradle infra error)
First seen : 2025-03-14
Total unique builds affected : 16
Pattern : Sporadic, 1–4 builds/month with gaps. Stable/low frequency.
Summary Table
#
Test
Builds Affected
First Seen
Trend
Reproduced
1
ClientYamlTestSuiteIT string profiler via global ordinals
350
2024-03
Stable/chronic
No
2
MixedClusterClientYamlTestSuiteIT string profiler via global ordinals
185
2024-03
Stable/chronic
Skipped
3
AwarenessAllocationIT testThreeZone...LoadAwareness
135
2024-08
Worsening
Inconclusive
4
RecoveryWhileUnderLoadIT testRecoverWhileUnderLoad...
117
2024-05
Worsening
Inconclusive
5
FullRollingRestartIT testFullRollingRestart...
104
2024-10
Worsening
No
6
AzureBlobStoreRepositoryTests testWriteRead
79
2024-04
Stable/chronic
Skipped
7
ConcurrentSeqNoVersioningIT testSeqNoCASLinearizability
66
2024-10
Stable/chronic
Yes
8
RestoreShallowSnapshotV2IT testRestoreShallowSnapshot...
50
2024-09
Stable/chronic
No
9
TelemetryMetricsEnabledSanityIT testGauge
30
2024-07
Worsening
No
10
WarmIndexSegmentReplicationIT testPrimaryStopped_ReplicaPromoted
16
2025-03
Stable/low freq
Inconclusive
Notes
WarmIndexSegmentReplicationIT.testReplicaHasDiffFilesThanPrimary (build 73690 , 9 unique builds affected, first seen 2025-03) was the 11th distinct failure and excluded from the top 10.
Tests marked "Inconclusive" failed due to Gradle test infrastructure errors (missing output files), not test assertion failures.
Tests marked "Skipped" could not be run locally due to environment requirements (BWC cluster versions, Docker).
The string profiler via global ordinals failures (Bump com.diffplug.spotless from 5.6.1 to 6.2.0 #1 and Bump com.diffplug.spotless from 5.6.1 to 6.2.1 #2 ) share the same root cause — a YAML REST test assertion that depends on segment topology.
Flaky test report: committed-code failures on 2026-04-01
Summary
11 test failures were detected across committed-code builds (Timer and Post Merge Action) in the past 24 hours. These map to 10 distinct failing tests (capped at 10 for this report). One test (ConcurrentSeqNoVersioningIT) reproduced locally with the original seed.
Failing Tests
1. ClientYamlTestSuiteIT —
string profiler via global ordinals2F166AF4E6F1B4F1field [profile.shards.0.aggregations.0.debug.segments_with_single_valued_ords] is not greater than [0]2. MixedClusterClientYamlTestSuiteIT —
string profiler via global ordinals6D4A15A66DCB92CFsegments_with_single_valued_ordsnot greater than 03. AwarenessAllocationIT —
testThreeZoneOneReplicaWithForceZoneValueAndLoadAwarenessD2E4E2BABCE53B0CExpected: <120> but: was <118>4. RecoveryWhileUnderLoadIT —
testRecoverWhileUnderLoadAllocateReplicasRelocatePrimariesTestE6978E901AAC624NoSuchElementException: No value present5. FullRollingRestartIT —
testFullRollingRestart_withNoRecoveryPayloadAndSource6E21A05A37A8A2E3replica shards haven't caught up with primary expected:<18> but was:<15>6. AzureBlobStoreRepositoryTests —
testWriteReadRepositoryVerificationException: path not accessible on cluster-manager node(BlobAlreadyExists/CONFLICT)7. ConcurrentSeqNoVersioningIT —
testSeqNoCASLinearizabilityC23880DF2C94CBC3ClusterHealthResponse has timed outC23880DF2C94CBC38. RestoreShallowSnapshotV2IT —
testRestoreShallowSnapshotIndexAfterSnapshotBD4077D675BFCA75timed out waiting for green state9. TelemetryMetricsEnabledSanityIT —
testGaugeFA09294D2AF3FDAAexpected:<2.0> but was:<3.0>10. WarmIndexSegmentReplicationIT —
testPrimaryStopped_ReplicaPromoted8644DC5BA7FFD3B5RefreshFailedEngineException: Refresh failed(CorruptIndexException: misplaced codec footer)Summary Table
string profiler via global ordinalsstring profiler via global ordinalstestThreeZone...LoadAwarenesstestRecoverWhileUnderLoad...testFullRollingRestart...testWriteReadtestSeqNoCASLinearizabilitytestRestoreShallowSnapshot...testGaugetestPrimaryStopped_ReplicaPromotedNotes
string profiler via global ordinalsfailures (Bump com.diffplug.spotless from 5.6.1 to 6.2.0 #1 and Bump com.diffplug.spotless from 5.6.1 to 6.2.1 #2) share the same root cause — a YAML REST test assertion that depends on segment topology.