[jira] [Created] (SPARK-17667) Make locking fine grained in YarnAllocator#enqueueGetLossReasonRequest

["Ashwin Shankar (JIRA)" <ji...@apache.org>]

Ashwin Shankar created SPARK-17667:
--------------------------------------

             Summary: Make locking fine grained in YarnAllocator#enqueueGetLossReasonRequest
                 Key: SPARK-17667
                 URL: https://issues.apache.org/jira/browse/SPARK-17667
             Project: Spark
          Issue Type: Bug
          Components: YARN
    Affects Versions: 2.0.0, 1.6.2
            Reporter: Ashwin Shankar


Following up on the discussion in SPARK-15725, one of the reason for AM hanging with dynamic allocation(DA) is the way locking is done in YarnAllocator. We noticed that when executors go down during the shrink phase of DA, AM gets locked up. On taking thread dump, we see threads trying to get loss for reason via YarnAllocator#enqueueGetLossReasonRequest, and they are all BLOCKED waiting for lock acquired by allocate call. This gets worse when the number of executors go down are in the thousands, and I've seen AM hang in the order of minutes. This jira is created to make the locking little more fine grained by remembering the executors that were killed via AM, and then serve the GetExecutorLossReason requests with that information.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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