[
https://issues.apache.org/jira/browse/CASSANDRA-1470?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12906330#action_12906330
]
Peter Schuller commented on CASSANDRA-1470:
-------------------------------------------
I do believe posix_fadvise() should work with DONTNEED, but it requires
fadvising() repeatedly after I/O. In addition for writes, one must fdatasync()
or fsync() first in order for a DONTNEED to have any effect. To what extent
this is an artifact of the Linux implementation and to what extent it is
supposed to work like that, I don't know.
(Not that I'm arguing for fadvise() over direct I/O, but FWIW.)
> use direct io for compaction
> ----------------------------
>
> Key: CASSANDRA-1470
> URL: https://issues.apache.org/jira/browse/CASSANDRA-1470
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Jonathan Ellis
> Fix For: 0.6.6
>
> Attachments: 1470-v2.txt, 1470.txt
>
>
> When compaction scans through a group of sstables, it forces the data in the
> os buffer cache being used for hot reads, which can have a dramatic negative
> effect on performance.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
