cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sylvain Lebresne (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-4885) Remove or rework per-row bloom filters
Date Tue, 19 Mar 2013 07:41:17 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-4885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13606144#comment-13606144
] 

Sylvain Lebresne commented on CASSANDRA-4885:
---------------------------------------------

bq. If we were writing Cassandra from scratch today

But we're not. I do think that providing an option and removing it is different from never
implementing it in the first place because in the former case some people may have start relying
on it. I also don't share your confidence that row-level BF are only ever useful in some niche
workloads that we don't care about (I do suspect that's largely true but I'm not totally confident).
Anyway, I can agree to disagree so do go on. After all, I don't disagree they should be disabled
by default, I just disagree that it's not worth keeping them as an option just in case we
had under-evaluated how useful they can be.

bq. you can come up with workloads that row cache benefits too.

Well, I happen to disagree about the removal of the row cache (until we have something better
that is), so I guess I'm coherent here.
                
> Remove or rework per-row bloom filters
> --------------------------------------
>
>                 Key: CASSANDRA-4885
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-4885
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core
>            Reporter: Jonathan Ellis
>            Assignee: Jason Brown
>             Fix For: 2.0
>
>         Attachments: 0001-CASSANRDA-4885-Remove-per-row-bloom-filter.patch, 0002-CASSANRDA-4885-update-test.patch,
4885-v1.patch
>
>
> Per-row bloom filters may be a misfeature.
> On small rows we don't create them.
> On large rows we essentially only do slice queries that can't take advantage of it.
> And on very large rows if we ever did deserialize it, the performance hit of doing so
would outweigh the benefit of skipping the actual read.

--
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

Mime
View raw message