aruraghuwanshi commented on PR #18:
URL: https://github.com/apache/druid-operator/pull/18#issuecomment-4540958922
> Also thinking on simpler terms, shouldn't the MM process support graceful
termination ? on the k8s we can just increase the terminationGracePeriods ?
Unfortunately Druid today doesn't have a built-in drain-on-shutdown for
MiddleManagers. When the MM process receives SIGTERM, the tasks actually end up
getting killed and any queries hitting those peons fail mid-execution.
The gracefulShutdownTimeout just controls how long the MM waits for the kill
to complete, so at least as of today, an external orchestrator is needed to
mark the MM as disabled and then call the POST
`/druid/indexer/v1/supervisor/{supervisorID}/taskGroups/handoff` which does not
kill the peons but marks those task groups ready for segment-handoff,
preventing any query failures due to MM restarts
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
