[
https://issues.apache.org/jira/browse/HBASE-3401?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Todd Lipcon updated HBASE-3401:
-------------------------------
Attachment: hbase-3401.txt
Here's a patch which adds a new QosPriority annotation which can be used to
prioritize things more easily right in the interface.
We can eventually extend this to be more generalized, but this should work for
now and make it more obvious when adding new RPCs how to prioritize them.
> Region IPC operations should be high priority
> ---------------------------------------------
>
> Key: HBASE-3401
> URL: https://issues.apache.org/jira/browse/HBASE-3401
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 0.90.1
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Fix For: 0.90.1
>
> Attachments: hbase-3401.txt
>
>
> I manufactured an imbalanced cluster so one region server had 300 regions and
> the others had very few. I then ran balancer while hitting the high-load
> region server with YCSB. I observed that the rate of load shedding was VERY
> slow since the closeRegion IPCs were getting stuck at the back of the IPC
> queue.
> All of these important master->RS RPC calls should be set to high priority.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
