cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stefania (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-10529) Channel.size() is costly, mutually exclusive, and on the critical path
Date Thu, 15 Oct 2015 04:26:05 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-10529?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14958281#comment-14958281
] 

Stefania commented on CASSANDRA-10529:
--------------------------------------

I've removed the assertion, see patch attached.

h5. CI

http://cassci.datastax.com/view/Dev/view/stef1927/job/stef1927-10529-3.0-dtest/
http://cassci.datastax.com/view/Dev/view/stef1927/job/stef1927-10529-3.0-testall/

h5. cstar comparison

http://cstar.datastax.com/graph?command=one_job&stats=b38f2330-72ec-11e5-8625-0256e416528f&metric=op_rate&operation=2_read&smoothing=1&show_aggregates=true&xmin=0&xmax=44&ymin=0&ymax=161867.2

http://cstar.datastax.com/graph?command=one_job&stats=9279e6e0-72f0-11e5-a177-0256e416528f&metric=op_rate&operation=2_read&smoothing=1&show_aggregates=true&xmin=0&xmax=43.01&ymin=0&ymax=167997.5

I think the gains must be below the variance of cstar or there must be an issue with the test
parameters or environment; the patched mmap performs worse in both cases but it shouldn't
since we just removed an assertion.

> Channel.size() is costly, mutually exclusive, and on the critical path
> ----------------------------------------------------------------------
>
>                 Key: CASSANDRA-10529
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-10529
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>            Reporter: Benedict
>            Assignee: Stefania
>             Fix For: 3.0.0 rc2
>
>
> [~stefania_alborghetti] mentioned this already on another ticket, but I have lost track
of exactly where. While benchmarking it became apparent this was a noticeable bottleneck for
small in-memory workloads with few files, especially with RF=1. We should probably fix this
soon, since it is trivial to do so, and the call is only to impose an assertion that our requested
length is less than the file size. It isn't possible to safely memoize a value anywhere we
can guarantee to be able to safely refer to it without some refactoring, so I suggest simply
removing the assertion for now.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message