Github user mwws commented on the pull request:
https://github.com/apache/spark/pull/8760#issuecomment-187548596
Besides from what @squito stated above, I have simplified the configuration
to resolve @kayousterhout 's concern. By default, exact same blacklist logic
will be applied as original. User can enable new blacklist logic by set
`spark.scheduler.blacklist.advancedStrategy=true` as an experimental feature.
Comparing with original logic, "Advanced Strategy" enable node blacklist: If
more than 1 executors on same node are in blacklist for given stage, we put all
executors on that node in blacklist for this stage. More details you can refer
to comments and [unit
test](https://github.com/mwws/spark/blob/SPARK-8426/core/src/test/scala/org/apache/spark/scheduler/BlacklistTrackerSuite.scala).
And the code has been rebased to latest master.
---
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]
