[
https://issues.apache.org/jira/browse/SPARK-4957?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kay Ousterhout closed SPARK-4957.
---------------------------------
Resolution: Won't Fix
Closing this as won't fix since it's been open (with no activity) for a while
> TaskScheduler: when no resources are available: backoff after # of tries and
> crash.
> -----------------------------------------------------------------------------------
>
> Key: SPARK-4957
> URL: https://issues.apache.org/jira/browse/SPARK-4957
> Project: Spark
> Issue Type: Improvement
> Components: Scheduler
> Affects Versions: 1.1.0
> Reporter: Nathan Bijnens
> Labels: scheduler
>
> Currently the TaskSchedulerImpl retries scheduling if there are no resources
> available. Unfortunately it keeps retrying with a small delay. It would make
> sense to throw an exception after a number of tries, instead of hanging
> indefinitely.
> https://github.com/apache/spark/blob/cb0eae3b78d7f6f56c0b9521ee48564a4967d3de/core/src/main/scala/org/apache/spark/scheduler/TaskSchedulerImpl.scala#L164-175
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
