cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <j...@apache.org>
Subject [jira] Updated: (CASSANDRA-863) Default cache levels should be explicit, not percentage
Date Wed, 10 Mar 2010 17:55:27 GMT

     [ https://issues.apache.org/jira/browse/CASSANDRA-863?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Jonathan Ellis updated CASSANDRA-863:
-------------------------------------

    Attachment: 863-v2.txt

v2 -- change default keys cached to 200k, move examples of overriding that in sample config
to CFs not used by stress test, and make Super1 BytesType so stress performance isn't artificially
worse than standard1 from the comparator

> Default cache levels should be explicit, not percentage
> -------------------------------------------------------
>
>                 Key: CASSANDRA-863
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-863
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core
>    Affects Versions: 0.6
>            Reporter: Brandon Williams
>            Assignee: Brandon Williams
>            Priority: Trivial
>             Fix For: 0.6
>
>         Attachments: 863-v2.txt, 863.txt
>
>
> Now that we can specify  explicit numbers for row and key caches, the default config
should use this.  The reason is that it's very easy to grow your keys and not realize a percentage
is going to get you eventually.  It's better to have a system that has a low cache hit rate
and keeps going than a system that encounters memory pressure from overpopulating its cache
and slows to a crawl or dies.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message