DanielLeens commented on issue #10993:
URL: https://github.com/apache/seatunnel/issues/10993#issuecomment-4794941714

   Thanks for the additional numbers.
   
   At this point, this still does not look like a confirmed scheduler bug yet. 
The important reason is that the stack trace in the issue body shows one 
concrete job failing during checkpoint because Kafka `InitProducerId` timed 
out, but that is different from proving why later submissions stay in `PENDING`.
   
   The most useful next step is to separate those two paths and capture the 
pending-job diagnostics for one specific stuck job on the same cluster state. 
Please provide:
   1. the result of `GET /pending-jobs?jobId=<one_pending_job_id>&limit=10`
   2. the result of `GET /system-monitoring-information` from the same time 
window
   3. the master log around that same pending job ID
   4. whether the same symptom still reproduces after moving from 2.3.11 to 
2.3.13+ and applying the concurrency tuning already suggested, especially 
`job-metrics-partition-count`
   
   Those four items will let us tell whether the job is waiting for slot 
assignment, blocked by coordinator pressure, or being affected by a different 
lifecycle problem.
   
   So for now, I would treat this as: worth continuing, but still needs one 
concrete pending-job diagnostic snapshot before we can call it a confirmed 
engine bug.


-- 
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]

Reply via email to