[
https://issues.apache.org/jira/browse/YARN-320?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Daryn Sharp updated YARN-320:
-----------------------------
Attachment: YARN-320.branch-23.patch
Proposed minimal change that fakes out the renewer. If the RM is talking to
itself (RPC UGI matches login UGI), then it will claim the renewer is renewing
the token.
This also has the benefit of allowing the client to request a RM token that
either it or a designated user can externally renew. Whereas today it's either
the RM renews or someone else, but not both.
It's been manually tested but I'll try to write unit tests and submit if this
approach is ok.
> RM should always be able to renew its own tokens
> ------------------------------------------------
>
> Key: YARN-320
> URL: https://issues.apache.org/jira/browse/YARN-320
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
> Affects Versions: 3.0.0, 2.0.3-alpha, 0.23.6
> Reporter: Daryn Sharp
> Assignee: Daryn Sharp
> Priority: Blocker
> Attachments: YARN-320.branch-23.patch
>
>
> YARN-280 introduced fast-fail for job submissions with bad tokens.
> Unfortunately, other stack components like oozie and customers are acquiring
> RM tokens with a hardcoded dummy renewer value. These jobs would fail after
> 24 hours because the RM token couldn't be renewed, but fast-fail is failing
> them immediately. The RM should always be able to renew its own tokens
> submitted with a job. The renewer field may continue to specify an external
> user who can renew.
--
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
