You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Jason Brown (JIRA)" <ji...@apache.org> on 2014/06/10 17:36:03 UTC
[jira] [Comment Edited] (CASSANDRA-6890) Standardize on a single
read path
[ https://issues.apache.org/jira/browse/CASSANDRA-6890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026475#comment-14026475 ]
Jason Brown edited comment on CASSANDRA-6890 at 6/10/14 3:35 PM:
-----------------------------------------------------------------
Hmm, are we sure we really want to do this before CASSANDRA-5863 is proven (or even implemented)? The v1 patch is simple enough, but we are taking a known performance hit against an unknown future payoff. How much simpler would removing the mmap path make implementing CASSANDRA-5863? I can't imagine it'd make that much of a difference?
Furthermore, if anyone is truly interested, I'll be talking about direct i/o (O_DIRECT) at the 2014 NGCC this week.
was (Author: jasobrown):
Hmm, are we sure we really want to do this before CASSANDRA-5863 is proven (or even implemented)? The v1 patch is simple enough, but we are taking a known performance hit against an unknown future payoff. How much simpler would removing the mmap path make implementing CASSANDRA-5863? I can't imagine it'd make that much of a difference?
Furthermore, if anyone is truly interested, I'll be talking about direct i/o (O_DIRECT) at the 2014 NGCC.
> Standardize on a single read path
> ---------------------------------
>
> Key: CASSANDRA-6890
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6890
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Joshua McKenzie
> Assignee: Joshua McKenzie
> Labels: performance
> Fix For: 3.0
>
> Attachments: 6890_v1.txt, mmap_gc.jpg, mmap_jstat.txt, mmap_perf.txt, nommap_gc.jpg, nommap_jstat.txt
>
>
> Since we actively unmap unreferenced SSTR's and also copy data out of those readers on the read path, the current memory mapped i/o is a lot of complexity for very little payoff. Clean out the mmapp'ed i/o on the read path.
--
This message was sent by Atlassian JIRA
(v6.2#6252)