Github user squito commented on the pull request:
https://github.com/apache/spark/pull/12951#issuecomment-220379391
Hi @njwhite, I'm not sure I see a strong need for this -- I posted a msg on
jira (as Kay had earlier). We should keep discussion about the feature in
general there, for archive / searchability.
In any case I did look at the code, so a couple of comments about that, if
we do decide we want this feature. Unless I'm missing something, it doesn't
seem like `maxShare` is being used anywhere (except for one trivial spot to
check its > 0). I was expecting a change to `FairSchedulingAlgorithm` at
least, though I think it might actually be a more complicated change to
`TaskSchedulerImpl` that is required.
Also to go along with that, we'd want new test cases demonstrating how
`maxShare` gets used.
---
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
