[
https://issues.apache.org/jira/browse/CASSANDRA-4292?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13437930#comment-13437930
]
Yuki Morishita commented on CASSANDRA-4292:
-------------------------------------------
Tests set up:
Server - 1 node, uses 4292-v4 patch and current trunk
Client - 3 nodes, each runs stress insert (modified to write to different CF)
for 5,000,000 keys with Leveled Compaction Strategy
I take 'iostat -x 1' from beginning of stress load to end of all compactions.
Result is uploaded here:
https://docs.google.com/spreadsheet/ccc?key=0AsVe14L_ijtkdC1tMXJ2d3RVLWxKVGxvWkgyVGFCWlE
>From above charts, you can see slight improvement in compaction speed(patched
>version ended fast). Also, patched version uses all disks evenly compared to
>trunk.
> Improve JBOD loadbalancing and reduce contention
> ------------------------------------------------
>
> Key: CASSANDRA-4292
> URL: https://issues.apache.org/jira/browse/CASSANDRA-4292
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Jonathan Ellis
> Assignee: Yuki Morishita
> Fix For: 1.2.0
>
> Attachments: 4292.txt, 4292-v2.txt, 4292-v3.txt, 4292-v4.txt
>
>
> As noted in CASSANDRA-809, we have a certain amount of flush (and compaction)
> threads, which mix and match disk volumes indiscriminately. It may be worth
> creating a tight thread -> disk affinity, to prevent unnecessary conflict at
> that level.
> OTOH as SSDs become more prevalent this becomes a non-issue. Unclear how
> much pain this actually causes in practice in the meantime.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
