nevzheng opened a new pull request, #14006: URL: https://github.com/apache/gravitino/pull/14006
### What changes were proposed in this pull request? - Logs every container start as `[CONTAINER-TIMING]` (pull, start, readiness); adds test log4j configs to the five modules whose logs dropped those lines. - `ContainerSuite.close()` resets all state; Hive containers are keyed by version; the PostgreSQL retry waits; the StarRocks image is pinned; an unused start method is removed. - Readiness checks poll with no initial delay; `cleanDisk()` runs only on real starts. - idp-basic starts its containers once per run; Ryuk is back on, without privileged mode. - Hive2, Flink and Spark ITs no longer close the shared suite. This PR includes the commits of the PRs below it. Review only its own commit: https://github.com/nevzheng/gravitino/compare/claude/13900-tc-01-design...claude/13900-tc-02-measure-lifecycle ### Why are the changes needed? See #14001 and the design doc's Background. Measured locally: Hive start −41 %, idp-basic −77 %, hive-metastore-common −36 %, catalog-hive −10 %, flink −16 %; a killed JVM's containers are removed in 15 s. Fix: #14001 ### Does this PR introduce _any_ user-facing change? No. Test code only. ### How was this patch tested? New unit tests for the reset, the Hive keying, the timing line and the PostgreSQL retry; Docker runs of idp-basic, hive-metastore-common, catalog-hive, flink-1.20 and spark-3.5 before and after (design doc, Appendix A). **Stack:** STACKLINE -- 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]
