[CASSANDRA-16371] JMH Reports in ci-cassandra.apache.org

• Type: Task
• Status: Resolved
• Priority: Low
• Resolution: Fixed
• Fix Version/s: 2.2.20, 3.0.25, 3.11.11, 4.0-rc1, 4.0
• Component/s: CI, Test/benchmark
• Change Category: Quality Assurance
• Complexity: Normal
• Platform: All
• Impacts: Docs

Source Control Link

https://github.com/apache/cassandra/commit/d19f35243733d211a81de149ef3b995fc7310854 https://github.com/apache/cassandra-builds/commit/8eff3855117b37d304711ce182edc516ea0fbd35

Test and Documentation Plan

Commit cassandra-builds patch first, manually trigger *-microbench jobs, validate known failures, commit cassandra patch.

• Assignee: Michael Semb Wever Assign to me
• Reporter: Michael Semb Wever

Authors

Michael Semb Wever

Reviewers

Tomasz Lasica

• Created: 29/Dec/20 13:35
• Updated: 08/Feb/21 10:40
• Resolved: 08/Feb/21 10:40

Description

Enable the jenkins jmh reporting plugin and add a jenkins job to run the `ant microbench` target.

Patches

• cassandra-builds: https://github.com/apache/cassandra-builds/compare/trunk...thelastpickle:mck/jenkins-microbench
• cassandra: https://github.com/apache/cassandra/compare/trunk...thelastpickle:mck/trunk_jmh_report

Example results are:

• CI JMH report: https://ci-cassandra.apache.org/job/Cassandra-devbranch-microbench/jdk=jdk_11_latest,label=cassandra/1/jmh-run-report/ (screenshot attached)

The value of these reports is limited, with different hardware jenkins agents and two executors on each agent. But visibility from the reports shows what is possible, how agents are different (and behave under saturation), and most importantly when jmh classes are broken (there are no unit tests on these benchmark tests).