[
https://issues.apache.org/jira/browse/CASSANDRA-20551?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17944869#comment-17944869
]
Jon Haddad edited comment on CASSANDRA-20551 at 4/15/25 11:47 PM:
------------------------------------------------------------------
Can confirm I'm seeing this in a CPU profile:
!screenshot-1.png|width=388,height=213!
was (Author: rustyrazorblade):
Can confirm I'm seeing this in a CPU profile:
!screenshot-1.png!
> ThreadLocalReadAheadBuffer shouldn't allocate on close if nothing is allocated
> ------------------------------------------------------------------------------
>
> Key: CASSANDRA-20551
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20551
> Project: Apache Cassandra
> Issue Type: Improvement
> Components: Local/Compaction
> Reporter: Jordan West
> Assignee: Jordan West
> Priority: Normal
> Attachments: screenshot-1.png
>
>
> Post the merge of CASSANDRA-15452 it was noticed that in
> {{ThreadLocalReadAheadBuffer#close()}} that we allocate a buffer even when
> not necessary. We can avoid this by not calling {{getBlock()}} and instead
> calling {{block()}} and returning early if the buffer is null.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
