[
https://issues.apache.org/jira/browse/CASSANDRA-4303?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13287349#comment-13287349
]
Daniel Doubleday commented on CASSANDRA-4303:
---------------------------------------------
Out of curiosity because I already started experimenting in this direction:
Has anybody benchmarked the option to keep bloom filters on disk. I did some
simple tests and could read around 6M random ints / sec from a paged-in
SegmentedFile. My line of thinking was: why not letting the os decide where the
memory should be allocated because its a tradeoff sst vs other data kept in mem
anyways.
I'm not close to a patch to do some real world testing so maybe you can save me
from the effort because someone else tried and learned that it does not make
sense.
> Compressed bloomfilters
> -----------------------
>
> Key: CASSANDRA-4303
> URL: https://issues.apache.org/jira/browse/CASSANDRA-4303
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Brandon Williams
> Fix For: 1.2
>
>
> Very commonly, people encountering an OOM need to increase their bloom filter
> false positive ratio to reduce memory pressure, since BFs tend to be the
> largest shareholder. It would make sense if we could alleviate the memory
> pressure from BFs with compression while maintaining the FP ratio (at the
> cost of a bit of cpu) that some users have come to expect. One possible
> implementation is at http://code.google.com/p/javaewah/
--
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
