Github user jerryshao commented on the pull request:
https://github.com/apache/spark/pull/8714#issuecomment-158859608
Hi @andrewor14 , from the code that `worker` never set the state to
`RUNNING`, currently when executor is started, `worker` will set the executor
state to `LAUNCHING` to report back to `master`. `Master` will notify
`AppClient` with `ExecutorAdded`, and `AppClient` will change the executor
state to `RUNNING` and report back to `master`. So simply, it is `AppClient`
who change the state to `RUNNING`, worker will only change to `LAUNCHING`.
So from my understanding of your description, that `worker` should set the
state to `RUNNING` in `ExecutorRunner#start()`, is this what you expected? Also
I think `AppClient` is not the role who should change the state of executor to
`RUNNING`, `AppClient` to my understand should only be notified with no right
to change, what do you think?
---
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]
