You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@myriad.apache.org by "John Yost (JIRA)" <ji...@apache.org> on 2016/06/15 19:53:09 UTC

[jira] [Created] (MYRIAD-222) Triage killable tasks to prevent retries on inaccessible nodes

John Yost created MYRIAD-222:
--------------------------------

             Summary: Triage killable tasks to prevent retries on inaccessible nodes
                 Key: MYRIAD-222
                 URL: https://issues.apache.org/jira/browse/MYRIAD-222
             Project: Myriad
          Issue Type: Improvement
            Reporter: John Yost
            Priority: Minor


When a Myriad task is marked killable an attempt is made to kill the task via the MyriadDriver, which delegates the call to the underlying MesosSchedulerDriver. In addition, the TaskTerminator daemon periodically attempts to kill any killable tasks that remain in the SchedulerState.

Darin suggested adding configurable logic to partition task ids pertaining to tasks for which multiple unsuccessful kill attempts have been made; this feature will prevent repeated attempts to kill tasks on inaccessible nodes.



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