Governance Meeting, March 10, 2025

Community
,
1

2025-03-10T17:30:17Z

Attendees :busts_in_silhouette:

Upcoming Calendar :tear_off_calendar:

  • Next weekly release: 2.501 Tuesday March 12, 2025
  • Next LTS: 2.492.3 - April 2, 2025 - Bruno Verachten is release lead
    • Release candidate March 19, 2025
    • Release April 2, 2025
  • LTS baseline selection: March 19, 2025
  • Next major events:
    • Linux Foundation Member Summit - Napa, CA, USA - March 18-20, 2025
    • cdCon and Open Source Summit North America - Denver, CO, USA - June 23-25, 2025

Agenda

News

Action Items

  • Basil create the attribution entries for the downloads page
  • Mark Waite submit Jenkins Contributor Summit summary blog post
  • Mark Waite submit quarterly CDF status report (was due 7 Mar 2025)
  • Mark Waite track reimbursements of all SPI funded attendees at Jenkins Contributor Summit
    • Payments confirmed by SPI for
      • Thorsten Scherler
      • Phillip Glanz
      • Valentin Delaye
      • Stefan Spieker
      • Jim Klimov
      • Daniel Krämer
      • Tim Jacomb
      • Alexander Brandes
    • Reiumbursement not yet requested for
      • Jan Faracik
  • Mark Waite propose further definition of sponsorship tiers

Community activity

  • Summary of Southern California Linux Expo (SCaLE 22x)
    • Alyssa Tong, Basil Crow, and Mark Waite ran the Jenkins booth
      • Good discussions with booth visitors
    • Scott Williams presented Scalable Open Source Data Science on Bare Metal Kubernetes
      • Jenkins automates the University of California Santa Barbara computer science classroom environments for data science
    • Mark Waite presented a 5 minute talk - misperceptions in open source contribution
  • Summary of Contributor Summit
    • Mark Waite preparing a blog post summary
      • Agenda was very busy, completed the goals, and had good working sessions
      • Blog post draft from Stefan Spieker on the community plugin maintainer inititative
      • Breakout sessions were a great idea, allow people to join the area of interest
        • Maintenance
        • UX SIG
        • Jenkins 3
      • Good to hear ideas of what people wanted to work on next
        • Better search (backend and frontend) - how to channel that interest into code
        • We’re good at small changes, but larger changes are much slower
        • Lots of motivation, even on larger projects, but need to find ways to channel the interest into results in a reasonable amount of time
        • Lots of motivation
    • Mark Waite using blog post summary as Jenkins project summary to CDF Technical Oversight Committee
    • Community plugin maintainers proposal for discussion in Jenkins developer mailing list
      • Java 25 is another project coming that will sweep across all the plugins
      • Jakarta EE 9 sweep has been much more smooth than the last sweep a year ago
        • Did not have to battle with outdated sources in top 250 plugins
  • Jenkins Content Security Policy project part 2
    • Project proposal submitted to Michael Winser of Alpha Omega
    • Project approval seems unlikely

Governance Topics

  • JFrog artifact storage reduction project (infra issue 4533)
    • JFrog sponsorship allows not more than 5 TB storage, we were over 10 TB storage
    • Darin Pope reduced our storage to less then 4 TB
      • Remove outdated artifacts from incrementals (saved 1 TB)
      • Dramatically reduced what we cache from Atlassian public
      • Other reductions as necessary
    • Refer to 2023 summary of previous bandwidth reduction project
  • Cloud expenses and plans
    • Azure (CDF paid)
      • August: $4.5k
      • September: $3.9k
      • October: $4.2k
      • November: $4.3k
      • December: $4.4k
      • January: $4.3k
      • February: $4.0k
    • Azure Sponsorship (Microsoft Credits) - $41k remaining, donation ends May 2025
      • August: $10.5k
      • September: $10.3k
      • October: $12.9k
      • November: $13k
      • December: $9.5k
      • January: $13.1k
      • February: $11.2
    • DigitalOcean - Remaining $15k (~5k consumed) until 02 January 2025
      • August $200
      • September: $158
      • October: $196
      • November: $146
      • December: $192
      • January: $219
      • February: $237
    • AWS:
      • CloudBees:
        • August: $6.3k
        • September: $6.3k
        • October: $6.4k
        • November: $3.9k
        • December: $540
        • January: $543
        • February: $550
      • Sponsored account
        • October: $178
        • November: $482
        • December: $600
        • January: $1400
        • February: $8.5k
  • Future of Gradle toolchain for plugin development
    • Gradle toolchain for plugin development is behind the Maven toolchain
      • Gradle JPI plugin was not usable with latest LTS until change from Basil
      • Basil converted a simple plugin as an example to use Gradle JPI
        • Steve Hill and others released a new version
      • Saw many other things out of sync, but not blockers
        • Not using the plugin BOM
        • Barely working now on the most recent LTS
        • Will need more work to bring it to parity with the Maven HPI plugin
      • No problem supporting it if it has feature parity
        • Problem when it does not have feature parity with Maven ecosystem
    • Oleg reviewed the JPI tools - possible Google Summer of Code project
      • Current Gradle JPI plugin needs additional work
      • Not impossible, just lots of work to do it
      • Does it make sense in the long term?
        • About 40 plugins use gradle, half those are deprecated
        • Gradle plugin is a recommended plugin but is not tested in plugin BOM
      • Would it make sense to convert gradle plugin to Maven
        • Need the gradle plugin to remain functional
    • Basil wrote an email message to the developer list
      • About a dozen plugins with more than 1000 installations
      • Gradle plugin has 230k installations, next closest is Multiple SCMs
      • Gradle build process is not widely used
      • Jenkins core team is not maintaining the gradle toolchain
        • Steve Hill and Rahul S are maintaining it for their use case
      • No desire to dictate a particular choice, but
    • New plugins won’t be hosted with Gradle
      • Considering the current state of Gradle JPI, Oleg supports that choice
      • Are there key contributions that ask to use Gradle with a new plugin
      • Gradle JPI gaps are a barrier to new plugin developers
    • Hosting checker specifically performs Maven checks that would need to be extended for Gradle
      • Validity checks in Maven would need to be developed for Gradle JPI plugin
    • Summary: If there is interest to contribute, could be useful and a good first step
      • Conventions plugin that represents a concept similar to the plugin parent pom
      • Had several follow-up questions in the Kotlin foundation
        • Put it in Kotlin because that is where Gradle participates in GSoC
      • GSoC project would be a difficult step, but not likely enough to complete the work
  • Next steps for a plugin adoption - Simon
    • Works for an open source company, customer needs to integrate their commits
    • Interested to help with the openstack cloud plugin
      • Read the mail from Oliver Gondza that he is looking for maintainers
      • Replied and pushed the pull request
      • Approval not needed, since the plugin is up for adoption
      • Last week adoption request, what’s next?
      • Hosting request looks correct, just need to process the request
        • Valid request, ready to process
        • 4 members of the hosting team, Tim, Alex, Adrien, Markus @jenkins-infra/hosting pings the whole team
          • Need to invite to the jenkinsci GitHub organization
      • Ask them to continue processing the request
  • Plugin BOM
    • Jenkins infrastructure is struggling to release the plugin BOM
      • What’s the primary goal of connecting the BOM release with the PCT?
        • Previously ran the entire suite on every pull request
        • Decided to save money by running only on weekly releases
          • Less than once a week we risk not detecting test failures
        • Need to run them enough to detect test failures
          • Balance the cost and the benefits
      • Infrastructure migration has not yet stabilized
        • If the tests don’t pass, then we either need to fix the test or the infra
2

FYI, my community.jenkins.io page is Profile - slide_o_mix - Jenkins