Github user squito commented on the issue:
https://github.com/apache/spark/pull/20640
lgtm
@IgorBerman can you cleanup the PR description a little? headers got
duplicated. And I'd reword a bit to something like
> This updates the Mesos scheduler to integrate with the common logic for
node-blacklisting across all cluster managers in BlacklistTracker.
Specifically, it removes a hardcoded MAX_SLAVE_FAILURES = 2 in
MesosCoarseGrainedSchedulerBackend, and uses the blacklist from the
BlacklistTracker, as Yarn does.
> This closes https://github.com/apache/spark/pull/17619
(the last "this closes" bit is useful for some tools we have, it will close
the other one when this is merged.)
thanks for doing this. I will probably leave this open for a bit if more
mesos users have thoughts
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
