[
https://issues.apache.org/jira/browse/CASSANDRA-20157?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Semb Wever updated CASSANDRA-20157:
-------------------------------------------
Attachment: Cassandra-5.0_352_timings.txt
> Timing info for each test split in Jenkinsfile
> ----------------------------------------------
>
> Key: CASSANDRA-20157
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20157
> Project: Apache Cassandra
> Issue Type: Improvement
> Components: CI
> Reporter: Michael Semb Wever
> Assignee: Michael Semb Wever
> Priority: Normal
> Fix For: 5.0.3, 5.1
>
> Attachments: Cassandra-5.0_352_timings.txt, Screenshot 2024-12-23 at
> 06.13.57.png, ci_summary_thelastpickle_mck-20157-5.0_138-1.html,
> consolelog__thelastpickle_mck-20157-5.0_138.txt.xz,
> results_details_thelastpickle_mck-20157-5.0_138.tar-1.xz
>
>
> Tuning the number of splits, and generally analysing where timeouts are (and
> will be) happening, is currently more difficult than it needs to be.
> Splits are created dynamically in the Jenkinsfile, with a fixed number of
> splits per test type. In trunk there's >900 splits. While the blue ocean ui
> can help highlight which splits have timed out (past the one hour per-split
> timeout) twice (each split will retry once), it remains difficult to scroll
> through the blue ocean screen, to identify splits that only timed out once,
> and to see which splits are heavier (taking longer) than their counterparts
> and the timing trends of splits over time.
> A simple log line reporting time taken for each split attempt solves all this.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
