[
https://issues.apache.org/jira/browse/CASSANDRA-9533?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14662908#comment-14662908
]
Benedict commented on CASSANDRA-9533:
-------------------------------------
If you're trickling one erase block every 10ms, I think most consumer grade
SSDs will still last years. So even that's probably not worth worrying about
It's also been shown in some workloads that having a delay may help spinning
disks, but my theoretical understanding of that is that it was unique to our
tool, and caused the work producer (stress) to phase its work production so
that there were fewer writes. This might be beneficial to some users as well,
but hard to say, and probably a minority.
I've committed to 2.1, 2.2, 3.0 and trunk.
> Make batch commitlog mode easier to tune
> ----------------------------------------
>
> Key: CASSANDRA-9533
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9533
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Assignee: Benedict
> Fix For: 3.x
>
>
> As discussed in CASSANDRA-9504, 2.1 changed commitlog_sync_batch_window_in_ms
> from a maximum time to wait between fsync to the minimum time, so one must be
> very careful to keep it small enough that most writers aren't kept waiting.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
