Started by GitHub push by ti-chi-bot[bot]00155681556843221432211.007489748916824fc74048f5ec88fa7fc68a155a0c3b6caf02b4e84fc74048f5ec88fa7fc68a155a0c3b6caf02b4e8origin/main4fc74048f5ec88fa7fc68a155a0c3b6caf02b4e84fc74048f5ec88fa7fc68a155a0c3b6caf02b4e8origin/main4fc74048f5ec88fa7fc68a155a0c3b6caf02b4e84fc74048f5ec88fa7fc68a155a0c3b6caf02b4e8origin/mainhttps://github.com/PingCAP-QE/ci.gitfalse#16824322141693seed #16821682falsefalse16823680279SUCCESS1778326286862https://do.pingcap.net/jenkins/job/seed/1682/default-hf41rpipelines/pingcap/tidb/latest/pull_unit_test_ddlv1.groovy4fc74048f5ec88fa7fc68a155a0c3b6caf02b4e81778326260000https://do.pingcap.net/jenkins/user/noreplynoreplynoreply@github.comci(tidb): align ddlv1 unit test reporting with unit test job (#4581) ### What changed - Increase `pull_unit_test_ddlv1` workspace from `150Gi` to `200Gi`, aligning it with `ghpr_unit_test`. - Capture `bazel_coverage_test_ddlargsv1` output into `bazel-test.log`. - Add the same test result archive/report flow used by the regular unit test job: - archive Bazel JUnit XML - archive `bazel-test.log` - run `analyze-go-test-from-bazel-output.sh` - send test case run report - archive `bazel-*.log` and `bazel-*.json` ### Why `pull_unit_test_ddlv1` has been hard to diagnose when Bazel reports timeout/failure summaries, because the job did not archive and parse test output like the regular unit test job. This change aligns the diagnostics path and removes the workspace size difference. Signed-off-by: lyb <yebin.li@pingcap.com> 2026-05-09 11:31:00 +00004fc74048f5ec88fa7fc68a155a0c3b6caf02b4e8ci(tidb): align ddlv1 unit test reporting with unit test job (#4581)editpipelines/pingcap/tidb/latest/pull_unit_test_ddlv1.groovygithttps://do.pingcap.net/jenkins/user/noreplynoreply