TestIntegrationSuite/TestNexusUpdateWorkflowOperation/Verify_concurrent_Updates_with_same_UpdateID_are_idempotent_and_finish (added in #2417, 2026-07-17) fails intermittently across platforms and Go versions on main:
| Run |
Commit |
Failing job(s) |
| 29769212673 |
f34dc3da (main) |
integ-test windows/oldstable, ubuntu/oldstable, macos-intel/stable |
| 29837385253 |
39018aef (main) |
integ-test-cache windows/stable |
| 29945839696 |
PR #2478 (contrib-only change; root test/ module does not link it) |
integ-test + integ-test-cache windows/stable |
The subtest times out after ~28–30s of its 30s budget, consistent with a race/timing sensitivity in the concurrent-updates-same-UpdateID path rather than an environment-specific break (no single platform pattern). Note main CI has not had a fully green run since at least 2026-07-17; other rotating flakes observed in the same window: TestResourceBasedSlotSupplierManyActs, TestAsyncOperationFromWorkflow_CancellationTypes/Abandon, TestWorkerHeartbeatBasic, TestPayloadSizeErrorActivityResult — but the NexusUpdateWorkflowOperation subtest is the most frequent.
TestIntegrationSuite/TestNexusUpdateWorkflowOperation/Verify_concurrent_Updates_with_same_UpdateID_are_idempotent_and_finish(added in #2417, 2026-07-17) fails intermittently across platforms and Go versions onmain:f34dc3da(main)39018aef(main)test/module does not link it)The subtest times out after ~28–30s of its 30s budget, consistent with a race/timing sensitivity in the concurrent-updates-same-UpdateID path rather than an environment-specific break (no single platform pattern). Note
mainCI has not had a fully green run since at least 2026-07-17; other rotating flakes observed in the same window:TestResourceBasedSlotSupplierManyActs,TestAsyncOperationFromWorkflow_CancellationTypes/Abandon,TestWorkerHeartbeatBasic,TestPayloadSizeErrorActivityResult— but the NexusUpdateWorkflowOperation subtest is the most frequent.