Github user jdesmet commented on the pull request:
https://github.com/apache/spark/pull/9095#issuecomment-215998052
From a user point of view the closure of this issue as-is is unacceptable.
I cannot understand why one would allow wrong job accounting for the executors
as reported in Yarn. This could affect the integrity of an entire cluster due
to over scheduling.
Part of the discussion goes about explaining how to fix it with a different
resource scheduler - of which I do not understand the details - but there was
no documentation to be found.
I am looking at a pretty big cluster for a pretty big company with a lot of
yarn scheduled jobs running on it - this worries me. It is pretty common to
have executors running with 32 vcores or more, and when running with that much
on one node - I have to be sure that yarn does not schedule anything else in.
---
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]
