matrixorigin / matrixorigin/matrixone
[CI] Ubuntu/x86 UT context timeout in TestUpdateIgnoreRepeatedAliasesAdvanceGreedily
- Dominant language
- Go
- Stars
- 1.9k
- Forks
- 311
- Avg merge
- 1d 3h
- Merged PRs (30d)
- 768
Description
## Summary
The Ubuntu/x86 UT job for PR #28987 failed in `pkg/embed.TestUpdateIgnoreRepeatedAliasesAdvanceGreedily` with an embedded-cluster control-plane deadline under the full-suite runner budget. The failure is recorded as a separate narrow CI issue because no open or recently closed MatrixOne issue covers this exact test name and signature.
## Source PR and failed checks
- PR: https://github.com/matrixorigin/matrixone/pull/28987
- Exact tested head: `0ed92b5105c8f0507b34caaade70ef1c19ede4bd`
- Workflow run: https://github.com/matrixorigin/matrixone/actions/runs/35086785091 (attempt 1)
- Failed Ubuntu/x86 UT job: https://github.com/matrixorigin/matrixone/actions/runs/35086785091/job/104763637515
- Derivative Required check: https://github.com/matrixorigin/matrixone/actions/runs/35086785091/job/104785368874
## Failure evidence
The job ran the repository UT command through `make ut`:
```text
go test -mod=readonly -vet=off -short -v -json -tags matrixone_test -p 2 -timeout 40m -race github.com/matrixorigin/matrixone/pkg/embed github.com/matrixorigin/matrixone/pkg/tests/arrowload github.com/matrixorigin/matrixone/pkg/tests/dml github.com/matrixorigin/matrixone/pkg/tests/issues/isolated github.com/matrixorigin/matrixone/pkg/tests/partition github.com/matrixorigin/matrixone/pkg/tests/shard github.com/matrixorigin/matrixone/pkg/tests/sqlintegration github.com/matrixorigin/matrixone/pkg/tests/testutils github.com/matrixorigin/matrixone/pkg/tests/upgrade
```
The complete job log and the untruncated UT diagnostic report show:
```text
pkg/embed/update_ignore_repeated_alias_test.go:44
Received unexpected error: context deadline exceeded
Messages: update ignore right_join_t b natural right join right_join_t a set b.v=?,a.u=? where a.id=1
--- FAIL: TestUpdateIgnoreRepeatedAliasesAdvanceGreedily (120.00s)
```
Immediately before the failure, the embedded cluster logged repeated HAKeeper/LogService proposal, checker-state, heartbeat and cluster-refresh timeouts. The same job also had unrelated Arrow LOAD and Data Branch deadline failures, and the suite was finally cancelled at its hard budget with `make: *** [Makefile:513: ut] Error 124`; no sort-rollup test failed.
## Independence and control evidence
Using `.agents/skills/mo-dev/scripts/mo-cgo-test` with the CI-equivalent race, tags, and CGo setup, the exact focused test passed 3/3 on the PR head:
```text
.agents/skills/mo-dev/scripts/mo-cgo-test -vet=off -short -v -json -tags matrixone_test -p 2 -timeout 40m -race -count=3 -run '^TestUpdateIgnoreRepeatedAliasesAdvanceGreedily$' ./pkg/embed
PR head: PASS 3/3 (38.65s, 27.62s, 21.52s)
```
The same command passed 3/3 on clean current `up/main` at `125d943786` in an isolated temporary worktree (42.58s, 24.13s, 27.23s). The temporary native-input symlinks and worktree were removed after the run.
The PR does not modify `pkg/embed`, the failing test, HAKeeper, LogService, embedded-cluster lifecycle, or the UT harness. The new planner path is guarded by `GroupBy.Rollup`; this failing UPDATE IGNORE workload does not construct a ROLLUP plan. The PR's sort-rollup planner, executor, compile, lifecycle-race, SCA, UT Coverage, Proxy BVT, Pessimistic BVT, shared-build and Coverage evidence all pass.
## Reproduction
Run the CI-equivalent focused command above on an Ubuntu/x86 MatrixOne UT worker while the normal package-level full suite is under runner load. The observed failure occurs when the embedded control plane exceeds the test's context deadline; the focused PR-head and clean-main controls complete successfully.
This issue records the evidence before the single permitted failed-jobs rerun of workflow run `35086785091`.
Contributor guide
Assessment
This issue has not been assessed yet.