Github user tgravescs commented on the pull request:
https://github.com/apache/spark/pull/2001#issuecomment-62046346
So the yarnClientSchedulerBackend already has an asyncMonitor to check is
the yarn side got shutdown and if it does it does an sc.stop. If the yarn
client mode driver side is shutdown badly such that that code isn't run, then I
could argue that Spark common code should handle something like this. It would
be no different from running in spark standalone mode and having the driver go
down. If you are in a shutdown hook inside of YarnClientClusterScheduler then
the driver is obviously already shutting down so what would calling sc.stop at
that point really mean?
Are you actually seeing a case where the spark driver isn't shutdown
(sc.stop) in yarn client mode? If so what are the conditions?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
