serhiy-bzhezytskyy commented on PR #4784: URL: https://github.com/apache/solr/pull/4784#issuecomment-5552110091
Beasting wouldn't tell you much here — with the new fail-fast removed the lock test still passes, so it isn't sensitive to this change. What the change widens is *when* the check runs: the old code only on `indexExists && firstTime && !passOnPreviousState`, the new one whenever the index exists, so reload and taking over a previous core are newly covered. I instrumented `initIndex` and ran the core-lifecycle suites: it fires 4x with `reload=true`, across `CoreAdminHandlerTest`, `TestCorePropertiesReload` and `TestReloadAndDeleteDocs`. Ten suites, 96 tests, green both on this branch and on `main` — no false lock failures in the newly-covered cases. The operator message isn't lost either: `SolrIndexWriter` rethrows `LockObtainFailedException` with the same "already locked ... another Solr server" text, now with the cause chained. Not demonstrated by a test: the race the old check had — it obtained the lock, released it, then opened the writer later. That window is what the 7.0 deprecation note was about. -- 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]
