[jira] [Commented] (AMBARI-2325) Agent remains in Queued status and then times out

["Siddharth Wagle (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AMBARI-2325?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13678407#comment-13678407 ] 

Siddharth Wagle commented on AMBARI-2325:
-----------------------------------------

+1 LGTM
                
> Agent remains in Queued status and then times out
> -------------------------------------------------
>
>                 Key: AMBARI-2325
>                 URL: https://issues.apache.org/jira/browse/AMBARI-2325
>             Project: Ambari
>          Issue Type: Bug
>          Components: controller
>    Affects Versions: 1.2.4
>            Reporter: Sumit Mohanty
>            Assignee: Sumit Mohanty
>             Fix For: 1.2.5
>
>         Attachments: AMBARI-2325.patch
>
>
> The issues seems to be as follows:
> Request A times out because some task (host-role-command) ran for more than 20 minutes. After 20 minutes BE times-out the task. However, agent fails to timeout or kill the task so it continues to wait till the task is completed.
> Meanwhile, server gets another request and schedules a set of tasks on the host. The tasks get scheduled (QUEUED) at the agent but the agent does not pick up the tasks for execution (it never goes to IN_PROGRESS). If the task stays long enough in Queued then eventually BE will time it out. So the tasks can go from QUEUED to TIMED_OUT.
> The agent has a bug where if attempt to send SIGTERM throws an exception, then it never send SIGKILL thus will end up waiting till the process (created due to task execution) is completed.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira