[
https://issues.apache.org/jira/browse/CASSANDRA-21460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18114041#comment-18114041
]
Jon Haddad commented on CASSANDRA-21460:
----------------------------------------
Had a cluster that was deliberately backed up with SSTables - using 100MB
target sstable size and 64MB compaction throughput with ~ 60 pending
compactions each. I switched to cursor compaction at 13:28 (marked by the red
line) and waited.
!Screenshot 2026-09-10 at 3.42.02 PM.png|width=825,height=148!
You can see pending compaction tasks complete a bit faster on db2, with them
being virtually eliminated when the iterator path still has ~70 on both nodes.
!Screenshot 2026-09-10 at 3.44.56 PM.png|width=816,height=365!
!Screenshot 2026-09-10 at 3.45.54 PM.png|width=270,height=213!
> Support BTI
> -----------
>
> Key: CASSANDRA-21460
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21460
> Project: Apache Cassandra
> Issue Type: Sub-task
> Components: Local/Compaction
> Reporter: Jon Haddad
> Assignee: Jon Haddad
> Priority: Normal
> Attachments: Screenshot 2026-09-10 at 2.17.44 PM.png, Screenshot
> 2026-09-10 at 2.17.58 PM.png, Screenshot 2026-09-10 at 3.42.02 PM.png,
> Screenshot 2026-09-10 at 3.44.56 PM.png, Screenshot 2026-09-10 at 3.45.54
> PM.png
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
