[
https://issues.apache.org/jira/browse/SAMZA-506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14276077#comment-14276077
]
Chris Riccomini commented on SAMZA-506:
---------------------------------------
Adding a shutdown hook to the SamzaContainer, and setting runLoop.shutdownNow
to true should make this work. The hook should also block until SamzaContainer
is fully shutdown (or a timeout occurs).
> Shutdown container on SIGTERM
> -----------------------------
>
> Key: SAMZA-506
> URL: https://issues.apache.org/jira/browse/SAMZA-506
> Project: Samza
> Issue Type: New Feature
> Components: container
> Affects Versions: 0.8.0
> Reporter: Zach Cox
> Fix For: 0.9.0
>
>
> Many runtime platforms (e.g. Mesos) will issue a SIGTERM to stop the Samza
> container process. In this case it would be good to try to cleanly shutdown
> the container, so that messages stop being processed, offsets and changelogs
> are committed, etc. This can be accomplished using TaskCoordinator.shutdown,
> but there currently is no glue between that and a SIGTERM or jvm shutdown
> hook.
> Here is the relevant mailing list thread:
> http://www.mail-archive.com/dev%40samza.incubator.apache.org/msg02246.html
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
