Github user IgorBerman commented on the issue:
https://github.com/apache/spark/pull/11207
@bbossy so my work around: ive disabled cleanup of external shuffle
service and removing shuffle files by cron that finds files that were not
accessed in last X hours.
---
Github user IgorBerman commented on the issue:
https://github.com/apache/spark/pull/11207
@bbossy thanks. I'm experiencing them too despite the merged PR. I think I
have some direction why it fails, here is my setup:
0. v2.2.0
1. Dynamic allocation is on
2. min executors
Github user bbossy commented on the issue:
https://github.com/apache/spark/pull/11207
@IgorBerman correct, #11272 was a follow up on this and got merged. I still
experience issues with the shuffle service on Mesos, but I haven't been able to
pin point the cause.
---
Github user IgorBerman commented on the issue:
https://github.com/apache/spark/pull/11207
Hi @bbossy
just to make sure this PR wasn't merged in favor to
https://github.com/apache/spark/pull/11272?
---
-
To
