[
https://issues.apache.org/jira/browse/CASSANDRA-11555?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15266417#comment-15266417
]
Benjamin Lerer commented on CASSANDRA-11555:
--------------------------------------------
I think that it is reasonable to have the cache size being: {{1/256th of the
the heap or 10MB, whichever is greater}} but in this case the comments in
{{Config}} and in {{DatabaseDescriptor}} are misleading. It also seems that the
code in {{DatabaseDescriptor}} is not accurate, unless I misunderstood
something. {code}Math.min(Math.max(1, (int) (Runtime.getRuntime().maxMemory()
/ 1024 / 1024 / 256)), 10){code} should be {code}Math.max((int)
(Runtime.getRuntime().maxMemory() / FileUtils.MB /256)), 10){code}
> Make prepared statement cache size configurable
> -----------------------------------------------
>
> Key: CASSANDRA-11555
> URL: https://issues.apache.org/jira/browse/CASSANDRA-11555
> Project: Cassandra
> Issue Type: Improvement
> Components: CQL
> Reporter: Robert Stupp
> Assignee: Robert Stupp
> Priority: Minor
> Labels: docs-impacting
> Fix For: 3.x
>
>
> The prepared statement caches in {{org.apache.cassandra.cql3.QueryProcessor}}
> are configured using the formula {{Runtime.getRuntime().maxMemory() / 256}}.
> Sometimes applications may need more than that. Proposal is to make that
> value configurable - probably also distinguish thrift and native CQL3 queries
> (new applications don't need the thrift stuff).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
