[
https://issues.apache.org/jira/browse/MESOS-2863?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14737540#comment-14737540
]
Vaibhav Khanduja commented on MESOS-2863:
-----------------------------------------
Appreciate your response here ...
The only reason I have understood for second message TASK_KILLED after
TASK_FINISHED, is when reaped is sleeping, shutdown received, which sends
SIGSTOP (os::killtree) making child send another SIGCHLD .. calling reaped
again (this weird, I have never seen this happening since as per POSIX standard
a signal is not delivered if the process is currently in the signal handler,
but in this case it is "waitpid") ... btw is the child still alive or may be
some other children.
explanation of flow your end would help ... or as fix, for whatever reason
second message go, I was planning to move the task sending code into a method.
The method maintains the state of the task and blocks TASK_KILLED after
TASK_FINISHED ..
> Command executor can send TASK_KILLED after TASK_FINISHED
> ---------------------------------------------------------
>
> Key: MESOS-2863
> URL: https://issues.apache.org/jira/browse/MESOS-2863
> Project: Mesos
> Issue Type: Bug
> Reporter: Vinod Kone
> Assignee: Vaibhav Khanduja
> Labels: newbie++
>
> Observed this while doing some tests in our test cluster.
> If the command executor gets a shutdown() (e.g., framework unregistered)
> after sending TASK_FINISHED but before exiting (there is a forced sleep), it
> could send a TASK_KILLED update to the slave.
> Ideally the command executor should not send multiple terminal updates.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
