You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@samza.apache.org by "Shanthoosh Venkataraman (JIRA)" <ji...@apache.org> on 2018/04/17 18:35:00 UTC

[jira] [Created] (SAMZA-1665) Fix race condition when stopping StreamProcessor.

Shanthoosh Venkataraman created SAMZA-1665:
----------------------------------------------

             Summary: Fix race condition when stopping StreamProcessor.
                 Key: SAMZA-1665
                 URL: https://issues.apache.org/jira/browse/SAMZA-1665
             Project: Samza
          Issue Type: Bug
            Reporter: Shanthoosh Venkataraman
            Assignee: Shanthoosh Venkataraman


When the user thread stops the streamProcessor and onNewJobModelExpired event is executed from debounce thread at the same time, we observe the following issue:

A. User thread stops the StreamProcessor

B. Before ZkJobCoordinator is shutdown, onNewJobModelExpired is executed from debounce thread(which spawns a new container). 

C. User thread stops the ZkJobCoordinator. 

After the StreamProcessor shutdown sequence returns,  container thread is alive and zkJobCoordinator is dead. This results in a orphaned container when stopping stream processor.  We were able reproduce this locally.



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