Caideyipi commented on PR #19037: URL: https://github.com/apache/iotdb/pull/19037#issuecomment-6485853886
Independent re-review of the full PR diff at `67387d27c0fb`, including one-way template path matching, shared-cluster isolation, and preservation of internal consensus Pipes. This round found one actionable test-isolation issue, fixed in `8df231b051cdacb0c7f2163591077a9f276d0c29`. - **P2: wait for asynchronous Pipe deletion before reusing a cluster.** The shared [`TestUtils.dropAllPipes`](https://github.com/apache/iotdb/blob/67387d27c0fb/integration-test/src/test/java/org/apache/iotdb/db/it/utils/TestUtils.java#L1161) returned after issuing `DROP PIPE`, and `IoTDBShowReceiversIT` repeated the same immediate cleanup. Replication from a previous test could therefore overlap database cleanup or the next test. The helper now polls `SHOW PIPES` with a bounded 60-second wait in both SQL dialects until only internal `__consensus` Pipes remain. [`IoTDBShowReceiversIT`](https://github.com/apache/iotdb/blob/8df231b051cdacb0c7f2163591077a9f276d0c29/integration-test/src/test/java/org/apache/iotdb/pipe/it/single/IoTDBShowReceiversIT.java#L95) now uses this shared helper. Verification: - `mvn spotless:apply -P with-integration-tests` and `git diff --check` passed. - The new `IoTDBShowReceiversIT` verification attempt under `MultiClusterIT1` did **not run the IT**. Approved remote hosts `.20` and `.19` were tried. The final attempt used exact source archives for IoTDB `8df231b051c` and TsFile `8f665d113dbb5257b901c528c31c8bc24da2a018`, JDK 17, clean source builds, sanitized Maven settings, and task-local Maven 3.9.11. Dependency preparation exited with code 1 before compilation because `com.gradle:develocity-maven-extension:2.6.0` could not be resolved; Maven's resolver record reported `Network is unreachable`. Local execution was not attempted because only approximately 2.8 GiB of Windows commit memory remained. This is not a passing compilation or IT result. - The previously recorded remote runs at `de2c7d730951` passed `PartialPathTest` (19 tests) and `ClusterTemplateManagerTest` (4 tests), including the new `testStartsWithOrEqual` and `testCheckTemplateSetInfo` methods. These new methods have not been run locally. The earlier 9-test `IoTDBShowReceiversIT` pass at that revision predates this round's cleanup fix and does not validate `8df231b051c`. - Subscription ITs remain **unverified with subscription support enabled**. The prior OSS attempt did not pass: `CommonConfig.SUBSCRIPTION_ENABLED=false`, and the OSS Failsafe configuration excludes subscription ITs. A corresponding subscription-enabled build is still required; no subscription pass is claimed. - This round's remote processes, reservations, source/tool directories, and runner records were cleaned up; no task process or marker leftovers remained. The remote head is `8df231b051cdacb0c7f2163591077a9f276d0c29`, and the PR remains **draft**. Because this round found an issue affecting test logic, the fix requires a fresh independent review in a new session. No second review round was performed here. -- 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]
