Github user sryza commented on the pull request:
https://github.com/apache/spark/pull/2504#issuecomment-74037430
I think SPARK-3172 is probably the more useful portion of this patch.
Ultimately, users want to be able to trace their shuffle resource consumption
back to the transformations they used. A single shuffle spill metric makes it
difficult to determine which of the stage boundaries bordering a task (and from
there, the transformation that triggered the stage boundary) is the culprit.
---
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]
