Github user JoshRosen commented on the pull request:
https://github.com/apache/spark/pull/3741#issuecomment-68029290
@jongyoul I think @tnachen is suggesting that Akka's use of threadpools
might mean that threads which have properly inherited the settings from
`runAsSparkUser` might send messages to Akka actors which end up being
processed in thread-pool threads that don't inherit those settings. So, the
indirection involved in Akka message processing might mean that the userinfo
isn't preserved properly.
Because of this, I think it makes sense to put the `runAsSparkUser` as
close to the actual task launching as possible. It looks like this PR does
that, though.
@tnachen Just to clarify, would you say this is still good to merge? It
seems fine to me, but I just wanted to have some discussion to better
understand how the fix works.
---
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]
