smengcl opened a new pull request, #11548: URL: https://github.com/apache/ozone/pull/11548
Generated-by: Claude Code (Opus 5.5) ## What changes were proposed in this pull request? `TestReconAndAdminContainerCLI` intermittently fails in `@BeforeAll` with `RocksDatabaseException: Rocks Database is closed` from `setupRatisKey`. After the full OM snapshot, Recon reprocesses into a staging DB and swaps it in. `ReconDBProvider#replaceStagedDb` closes the current DB before `ReconContainerMetadataManager#reinitialize` points the tables at the new one, so a read in that window fails. `setupRatisKey` polled `getKeyCountForContainer` and rethrew any `IOException` as `RuntimeException`, failing the whole class instead of retrying. This patch reuses the existing `ReconOmMetaManagerTestUtils#waitUntilReconKeyCounts`, which treats such `IOException` as not ready yet and keeps polling. ## What is the link to the Apache JIRA https://issues.apache.org/jira/browse/HDDS-17037 ## How was this patch tested? - `mvn -pl :ozone-integration-test-recon -am test -Dtest=TestReconAndAdminContainerCLI`: passed locally. - The fix removes the failure mode by construction: a transient closed DB read during the staging swap no longer aborts the wait. One green run alone does not prove non-flakiness, hence the repeated run below. - 10x10 [flaky-test-check](https://github.com/smengcl/hadoop-ozone/actions/runs/38679633466) on this branch: in progress 🤖 Generated with [Claude Code](https://claude.com/claude-code) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
