Github user andrewor14 commented on the pull request:
https://github.com/apache/spark/pull/3783#issuecomment-69406817
The other thing is what to do about executors that should be removed but
were not because of the lower bound. On a high level, we should not forget that
these executors were once idle. We can do this by maintaining a set of such
executors, and marking these as idle as soon as new executors join, for
instance. (We could also do the `-1` thing I suggested earlier, but I think
that solution is slightly more complicated in terms of code.)
---
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]
