spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Miklos Szurap (JIRA)" <j...@apache.org>
Subject [jira] [Created] (SPARK-24449) ApplicationMaster reporter thread failure counter is not effective
Date Fri, 01 Jun 2018 09:49:00 GMT
Miklos Szurap created SPARK-24449:
-------------------------------------

             Summary: ApplicationMaster reporter thread failure counter is not effective
                 Key: SPARK-24449
                 URL: https://issues.apache.org/jira/browse/SPARK-24449
             Project: Spark
          Issue Type: Improvement
          Components: Spark Core
    Affects Versions: 2.3.0
            Reporter: Miklos Szurap


The ApplicationMaster's [failureCount|https://github.com/apache/spark/blob/master/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/ApplicationMaster.scala#L522] gets
reset on every while loop. This makes reporterMaxFailures / MAX_REPORTER_THREAD_FAILURES
({{spark.yarn.scheduler.reporterThread.maxFailures)}} ineffective.

The failureCount was intended to be used to fail the application if there were [non-fatal
Throwables|https://www.scala-lang.org/api/current/scala/util/control/NonFatal$.html] more
than reporterMaxFailures ( {{spark.yarn.scheduler.reporterThread.maxFailures}} ) times.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message