Github user tgravescs commented on a diff in the pull request:
https://github.com/apache/spark/pull/14079#discussion_r77893351
--- Diff:
core/src/test/scala/org/apache/spark/scheduler/BlacklistIntegrationSuite.scala
---
@@ -51,37 +54,68 @@ class BlacklistIntegrationSuite extends
SchedulerIntegrationSuite[MultiExecutorM
assertDataStructuresEmpty(noFailure = false)
}
- // even with the blacklist turned on, if maxTaskFailures is not more
than the number
- // of executors on the bad node, then locality preferences will lead to
us cycling through
- // the executors on the bad node, and still failing the job
+ // even with the blacklist turned on, bad configs can lead to job
failure. To survive one
+ // bad node, you need to make sure that
+ // maxTaskFailures > min(spark.blacklist.maxTaskFailuresPerNode,
nExecutorsPerHost)
--- End diff --
comment out of date spark.blacklist.maxTaskFailuresPerNode doesn't exist.
---
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
