SEZ9 commented on PR #13154: URL: https://github.com/apache/seatunnel/pull/13154#issuecomment-6687541310
CI status, since the red mark needs context. **First failure** was mine: spotless disagreed with my javadoc line wrapping in `SubPlan.java`. Fixed with `mvn spotless:apply` in `3fdac5365` — comment rewrapping only, no logic touched. **Second failure is not from this change.** `unit-test (17, windows-latest)` and `all-connectors-it-5` both died on a Maven Central rate limit: ``` Failed to execute goal on project seatunnel-translation-spark-common: Could not transfer artifact org.apache.spark:spark-streaming_2.12:pom:3.3.0 from/to central (https://repo.maven.apache.org/maven2): ... status: 429 [ERROR] you can resume the build with the command [ERROR] mvn <args> -rf :seatunnel-translation-spark-common ``` I checked for test failures before concluding that — grepping the full 63k-line log for `Failures: [1-9]` and `Errors: [1-9]` gives zero hits, and `SeaTunnel : Tools : CI : Java` passed, so the code-style gate is clear. One thing I want to be precise about rather than let the log imply otherwise: `seatunnel-translation-spark-common` sits **before** `seatunnel-engine-server` in the reactor, so the 429 skipped everything after it: ``` [INFO] SeaTunnel : Engine : Server ........................ SKIPPED ``` So this run did not compile or test the change either. Combined with the first failure stopping at code style, **the Java in this PR still has no verification behind it** — which is what I flagged in the description, and it is still true rather than resolved. I cannot build locally: Maven here cannot reach `repo.maven.apache.org` (`maven-site-plugin` and `guava` both fail to resolve). So I have pushed an empty commit to request a fresh run, since 429 is transient. If it reproduces I will look for another way to get the module compiled rather than keep re-triggering. What I have verified by reading: the capture point sits inside the existing `synchronized (restoreLock)`; `reset()` does not touch `errorByPhysicalVertex`, so the value is still present when captured; the failing vertex sets it through `compareAndSet` before the pipeline reaches the restore decision; `ConcurrentLinkedQueue.size()` is O(n) with n capped at 16; and `getRestoreCauses()` returns a copy. -- 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]
