[
https://issues.apache.org/jira/browse/YARN-9303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Prabhu Joseph updated YARN-9303:
--------------------------------
Description:
Currently Timeline Service Storage tables uses splits based on username.
ApplicationTable rowkey starts with ClusterId, AppToFlowTable rowkey starts
with AppId won;t work well with username splits. The Hbase write will hotspot
to one RegionServer. Need to optimize the pre split keys for each tables based
on the RowKey.
http://archive.cloudera.com/cdh5/cdh/5/hbase-0.98.6-cdh5.3.8/book/rowkey.design.html
was:Currently Timeline Service Storage tables uses splits based on username.
ApplicationTable rowkey starts with ClusterId, AppToFlowTable rowkey starts
with AppId won;t work well with username splits. The Hbase write will hotspot
to one RegionServer. Need to optimize the pre split keys for each tables based
on the RowKey.
> Optimizing Pre Split keys for Hbase Timeline Service Storage tables
> -------------------------------------------------------------------
>
> Key: YARN-9303
> URL: https://issues.apache.org/jira/browse/YARN-9303
> Project: Hadoop YARN
> Issue Type: Bug
> Affects Versions: 3.1.2
> Reporter: Prabhu Joseph
> Assignee: Prabhu Joseph
> Priority: Major
>
> Currently Timeline Service Storage tables uses splits based on username.
> ApplicationTable rowkey starts with ClusterId, AppToFlowTable rowkey starts
> with AppId won;t work well with username splits. The Hbase write will hotspot
> to one RegionServer. Need to optimize the pre split keys for each tables
> based on the RowKey.
> http://archive.cloudera.com/cdh5/cdh/5/hbase-0.98.6-cdh5.3.8/book/rowkey.design.html
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
