[
https://issues.apache.org/jira/browse/CASSANDRA-20296?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17929226#comment-17929226
]
Jon Haddad edited comment on CASSANDRA-20296 at 2/21/25 6:32 PM:
-----------------------------------------------------------------
Well, for the sake of argument I'll assume the server is busy. If we're not
getting a lot of flushes, that means we're not doing a lot of writes, and the
workload is read heavy. In that case, we'd actually be better off with an even
bigger new gen, because reads do a TON of allocation.
There's no downside, that I've found, in any workload, to 50% new gen.
was (Author: rustyrazorblade):
Well, for the sake of argument I'll assume the server is busy. If we're not
getting a lot of flushes, that means we're not doing a lot of writes, and the
workload is read heavy. In that case, we'd actually be better off with an even
bigger new gen, because reads do a TON of allocation.
> Heap and GC jvm flags improvements
> ----------------------------------
>
> Key: CASSANDRA-20296
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20296
> Project: Apache Cassandra
> Issue Type: Improvement
> Components: Local/Config
> Reporter: Michael Semb Wever
> Assignee: Michael Semb Wever
> Priority: Normal
> Fix For: 5.0.x, 5.x
>
> Attachments: ci_summary_thelastpickle-mck-20296-5.0_155.html,
> results_details_thelastpickle-mck-20296-5.0_155.tar.xz
>
>
> A few improvements can be made to the configuration files:
> * conf/cassandra-env.sh
> * conf/jvm-server.options
> * conf/jvm11-server.options
> * conf/jvm17-server.options
> G1 can rely on `-XX:G1NewSizePercent=50` to always floor the young
> generation's size to 50% of the heap. (In production this can be raised to
> 66% for optimal performance.)
> Fixing the young generation size with `-Xmn` is only applicable to CMS,
> so move it to the appropriate section.
> When using G1, we can automatically set (the default)
> `-XX:ParallelGCThreads` and `-XX:ConcGCThreads` to the number of system cpu
> cores.
> The auto-generated (default) heap size is now half the server's physical
> RAM, capped at 16G for CMS and 31G for G1.
> Increase MaxTenuringThreshold from 1 to 2, as evidence has shown it has
> no cost (over values of zero or one), but can have significant benefits in
> keeping objects in the young generation. While, values above 2 don't have
> any noticeable benefit.
> Always set CASSANDRA_HEAPDUMP_DIR, default it to $CASSANDRA_LOG_DIR, to
> avoid hprof filling up unexpected disk volumes. Assumption here is that the
> logs directory is large enough to handle these dumps, and/or operators are
> monitoring these directories more than other randon/unknown directories.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
