[
https://issues.apache.org/jira/browse/CASSANDRA-4310?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yuki Morishita updated CASSANDRA-4310:
--------------------------------------
Attachment: 4310-v3.txt
V3 attached.
bq. Shouldn't "skip submitBackground" logic be "CF is currently compacting, AND
no more idle threads?"
Yes. Fixed this.
bq. Looks like compactingCF.remove(cfs); could be moved to the topmost finally
block of run()
Done.
bq. We already track compacting sstables in DataTracker.View; seems like we
shouldn't need to duplicate this in LeveledManifest. (I note that getting this
correct in DT took some effort so I am worried on that level as well as
abstract code purity.) If the problem is that getNextBackgroundTask is called
multiple times before CompactionManager officially marks the sstables as
in-progress, my preferred solution would be to move the in-progress code
earlier.
I modified the code to mark sstables compacting before returning compaction
task so that I can use DataTracker.View's compacting property. If nothing can
be marked, then no task is returned.
I also cleaned up no longer used codes.
> Multiple independent Level Compactions in Parallel
> --------------------------------------------------
>
> Key: CASSANDRA-4310
> URL: https://issues.apache.org/jira/browse/CASSANDRA-4310
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Affects Versions: 1.0.0
> Reporter: sankalp kohli
> Assignee: Yuki Morishita
> Labels: compaction, features, leveled, performance, ssd
> Fix For: 1.2.1
>
> Attachments: 4310.txt, 4310-v2.txt, 4310-v3.txt
>
>
> Problem: If you are inserting data into cassandra and level compaction cannot
> catchup, you will create lot of files in L0.
> Here is a solution which will help here and also increase the performance of
> level compaction.
> We can do many compactions in parallel for unrelated data.
> 1) For no over lapping levels. Ex: when L0 stable is compacting with L1, we
> can do compactions in other levels like L2 and L3 if they are eligible.
> 2) We can also do compactions with files in L1 which are not participating in
> L0 compactions.
> This is specially useful if you are using SSD and is not bottlenecked by IO.
> I am seeing this issue in my cluster. The compactions pending are more than
> 50k and the disk usage is not that much(I am using SSD).
> I am doing multithreaded to true and also not throttling the IO by putting
> the value as 0.
>
--
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
