JoegenUSTC commented on PR #11355: URL: https://github.com/apache/gravitino/pull/11355#issuecomment-4611071228
> > > The CI fails. Can you take time to fix it? @JoegenUSTC > > > @yuqi1129 > > > Thanks for flagging this! I've fixed the CI failure in commit `11fbb32`. > > > > > > The root cause was that `TestHive2HMSWithKerberosNoImpersonation.startHiveContainer()` was redundantly constructing a second `HiveClientFactory` after `super.startHiveContainer()`. Since `TestHive2HMSWithKerberos.startHiveContainer()` already calls `createHiveProperties()` via polymorphism, the overridden `createHiveProperties()` (with `impersonation-enable=false`) was already in effect for the initial `hiveClient` creation. The redundant second construction failed in the CI Kerberos environment, leaving `hiveClient=null` and causing NPE in all test methods. > > The fix removes the `startHiveContainer()` override entirely — only `createHiveProperties()` needs to be overridden. > > All tests pass locally (`./gradlew :catalogs:hive-metastore-common:test -PskipITs`). Could you please approve the CI workflows to run? Thank you! > > Done. @yuqi1129 All three CI failures are transient infrastructure issues, unrelated to this PR's changes: 1. **`catalog-fileset:test`** — Job was canceled by the CI runner (timeout), not a test failure. 2. **`catalog-glue:test`** — Docker image pull timeout (`motoserver/moto:5.1.4` not fetched within 2 minutes); `trino-connector-435-439:compileJava` — JDK toolchain download timeout (`SocketTimeoutException`). Both are network issues on the CI runner. 3. **`trino-connector-473-478`** — Job was canceled by the CI runner (timeout), not a test failure. Could you please re-run the failed jobs? Thank you! -- 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]
