Github user vanzin commented on the issue:
https://github.com/apache/spark/pull/16345
Hmm, I see... the worker needs to listen for RPC connections also, because
executors connect to it for "fate sharing"... so that means that when it
connects to the master, there are two TCP channels involved instead of one for
"client mode" RpcEnv.
But, with your change, aren't you losing the timeout for the register RPC?
Basically the `case Failure(e)` in the code you're deleting from Worker.scala.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
