[
https://issues.apache.org/jira/browse/CASSANDRA-7040?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13969726#comment-13969726
]
Jason Brown commented on CASSANDRA-7040:
----------------------------------------
Martin Thompson mentions batching IO events in a talk at the react conf 2014:
https://www.youtube.com/watch?v=4dfk3ucthN8 . The idea seems reasonable but I
haven't investigated it yet.
bq. that may touch the disks
Yeah, the key word here is *may*. You could add in helpers like mincore (and
row cache) to help inform you if you have nothing in memory and that you'll be
going to disk.
> Replace read/write stage with per-disk access coordination
> ----------------------------------------------------------
>
> Key: CASSANDRA-7040
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7040
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Benedict
> Labels: performance
> Fix For: 3.0
>
>
> As discussed in CASSANDRA-6995, current coordination of access to disk is
> suboptimal: instead of ensuring disk accesses alone are coordinated, we
> instead coordinate at the level of operations that may touch the disks,
> ensuring only so many are proceeding at once. As such, tuning is difficult,
> and we incur unnecessary delays for operations that would not touch the
> disk(s).
> Ideally we would instead simply use a shared coordination primitive to gate
> access to the disk when we perform a rebuffer. This work would dovetail very
> nicely with any work in CASSANDRA-5863, as we could prevent any blocking or
> context switching for data that we know to be cached. It also, as far as I
> can tell, obviates the need for CASSANDRA-5239.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
