DanielLeens commented on issue #13319: URL: https://github.com/apache/seatunnel/issues/13319#issuecomment-6983639774
I reproduced the `add(null)` failure and incomplete tracker at the actual thread-share submission method on `89f91a89af15f33d97a942d8b65f4fc1b0c80bf5` (Java 11.0.27). Steps: 1. Prepare a `TaskExecutionService` fixture with a real `LinkedBlockingDeque` and two task doubles; set the tracker completion latch to 2. 2. Make the first task's `init()` throw an injected runtime exception. The second task increments a counter if initialized. 3. Invoke actual `submitThreadShareTask(tracker, tasks)`. 4. Observe `NullPointerException` from `LinkedBlockingDeque.add(null)`, reached through `TaskExecutionService.java:391`. 5. Verify the remaining latch is **1**, the second task's init count is **0**, and the tracker's completion future is still incomplete. [Executable Java probe and exact build/run commands](https://github.com/apache/seatunnel/issues/13318#issuecomment-6983607036): see `shareInit()` in `TaskIssueProbes.java`. The fixture uses reflection/Unsafe to initialize the narrow submission path and a real cancellation executor (shut down afterward). This confirms the submission/tracker failure signature. It does **not** exercise deployment-context publication, a complete redeployment, or prove a measured context/resource leak in a running worker. -- 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]
