Github user jerryshao commented on the issue:
https://github.com/apache/spark/pull/19812
Does this failure ". For some reason, all of the 3 executors failed. "
happened during task running or before task submission? Besides, if you're
running on yarn, yarn will bring new executors to meet the requirement, also
`ExecutorAllocationManager` will be notified with executor lost/register. Can
you please tell us how to reproduce your scenario?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]