[
https://issues.apache.org/jira/browse/YARN-1055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13740474#comment-13740474
]
Bikas Saha commented on YARN-1055:
----------------------------------
Why does the launcher not retry the action? Is there a jira in OOZIE to make it
work properly in such cases by doing its own book-keeping? Isnt it more correct
to fix OOZIE instead of adding a workaround config in YARN?
Is the current situation acceptable as a known short term bug? From what I see
nothing wrong will happen functionally/practically. In infrequent cases of the
action-AM node crashing, the pipeline would have to be restarted. We have a
design for work-preserving RM restart that can be completed post beta. This
will remove the need to restart AM's. Given that, I am really averse to adding
a short term work around API in AppSubmissionContext that will have to be
maintained till YARN-3.0 comes out because we are guaranteeing API's post beta.
> Handle app recovery differently for AM failures and RM restart
> --------------------------------------------------------------
>
> Key: YARN-1055
> URL: https://issues.apache.org/jira/browse/YARN-1055
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
> Affects Versions: 2.1.0-beta
> Reporter: Karthik Kambatla
>
> Ideally, we would like to tolerate container, AM, RM failures. App recovery
> for AM and RM currently relies on the max-attempts config; tolerating AM
> failures requires it to be > 1 and tolerating RM failure/restart requires it
> to be = 1.
> We should handle these two differently, with two separate configs.
--
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
