[
https://issues.apache.org/jira/browse/YARN-11741?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Haitian updated YARN-11741:
---------------------------
Affects Version/s: 3.4.1
4.0.0
3.5.0
> Follow-up on ResourceManager quit due to ApplicationStateData exceeding znode
> limit
> -----------------------------------------------------------------------------------
>
> Key: YARN-11741
> URL: https://issues.apache.org/jira/browse/YARN-11741
> Project: Hadoop YARN
> Issue Type: Bug
> Affects Versions: 4.0.0, 3.3.6, 3.4.1, 3.5.0
> Reporter: Haitian
> Priority: Major
>
> In YARN-5006 the issue "ResourceManager quit due to ApplicationStateData
> exceeding znode limit" has been addressed , specifically in the
> storeApplicationStateInternal() method of the ZKRMStateStore class. However,
> the methods updateApplicationStateInternal(),
> storeApplicationAttemptStateInternal(), and
> updateApplicationAttemptStateInternal() also need to incorporate checks for
> the ZNode size limit in ZKRMStateStore class. This will help prevent
> scenarios where excessive data written to a single ZNode causes the
> ResourceManager (RM) and ZooKeeper (ZK) services to become unavailable.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
