Github user harishreedharan commented on the pull request:
https://github.com/apache/spark/pull/6508#issuecomment-106944232
I will update this to add the timeout parameter. So I will update this to
check the expired executor (as is now) and check if there are cached blocks -
if yes, change the timeout to Long.MAX - so we don't end up sending RPCs every
100ms once we discover that it has cached blocks. @andrewor14, @sryza - that
should take care of the issue.
(Note that there is a slight race condition in this whole idea - if a new
block gets cached after we do the check, then we will remove the executor -
this can be fixed only with some form of distributed locking).
---
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]
