cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benedict (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (CASSANDRA-6838) FileCacheService dramatically overcounting its memoryUsage
Date Wed, 12 Mar 2014 00:54:43 GMT

     [ https://issues.apache.org/jira/browse/CASSANDRA-6838?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Benedict updated CASSANDRA-6838:
--------------------------------

    Attachment: 6838.txt

> FileCacheService dramatically overcounting its memoryUsage
> ----------------------------------------------------------
>
>                 Key: CASSANDRA-6838
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-6838
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>            Reporter: Benedict
>            Assignee: Benedict
>              Labels: performance
>             Fix For: 2.1 beta2
>
>         Attachments: 6838.txt
>
>
> On investigating why I was seeing dramatically worse performance for counter updates
over prepared CQL3 statements compred to unprepared CQL2 statements, I stumbled upon a bug
in FileCacheService wherein, on returning a cached reader back to the pool, its memory is
counted again towards the total memory usage, but is not matched by a decrement when checked
out. So we effectively are probably not caching readers most of the time.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message