Github user JoshRosen commented on the pull request:
https://github.com/apache/spark/pull/2250#issuecomment-56281680
I feel strongly that we should use the same application ID to refer to the
application in every context, since creating a different id based off of
System.currentTimeMillis could be very confusing for users. As a user, I'd
like to be able to grep logs / metrics / web UIs for my application data using
one application id; displaying some other unique but random value is confusing
because I have to compare timestamps, etc. to correlate the ids.
This is tricky, though, since we have a "chicken and egg" initialization
problem, as you've described. I like the approach that you've suggested in
#2432, so I'm going to continue review over there. Feel free to leave this PR
open, though, so that it shows up in our [PR
dashboard](https://spark-prs.appspot.com) and invites discussion; it will be
automatically closed if I merge your other PR.
---
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
