Re: [jakartaee-spec-project-leads] Getting started with CI/Jenkins - Jak

[Bill Shannon <bill.shannon@xxxxxxxxxx>]

    One of the challenges we currently face is that, because of
    permissions, we can't just say look at project X for an example,
    they did it right.
    
    If you create a Jenkins Pipeline job, then the Jenkinsfile in the
    repo drives the release, along with some shell scripts stored in the
    repo.  Many of our repos have this set up, so you can look at (e.g.)
    the "mail" repo, which I know something about, as an example.
    
    Look at the GitHub
      webhook configuration page in our wiki.  You might need to
    file bugzilla web tasks to get some of the steps done; I had admin
    privileges when I wrote that wiki page.
    
    If the webhook triggers the pipeline job, it will build the project
    and run the TCK on the project.
    
    You can set up another webhook that points at the TCK repo that will
    cause it to run the TCK against a stable version of the
    implementation project, which is good for testing changes to the
    TCK.
    
    When it gets time to release the project, whether a SNAPSHOT
    release, a milestone release, or a final release, most people are
    using the scripts describe in the wiki https://wiki.eclipse.org/JakartaEE_New_Infra_Release_Job
    and https://wiki.eclipse.org/EE4J_Maven_Publishing
    (see script at the bottom).  These scripts are always invoked
    manually.
    
    That should be enough to get you started.  Let me know if you have
    more questions.

Scott Kurz wrote on 3/17/20 7:58 AM:

Hi, I'm looking for pointers or first steps for getting the Jakarta Batch started using some of the CI / Jenkins pipelines used by other projects.

To date, I've been releasing the batch-api project manually via my own personal credentials, not via CI credentials. Haven't had time to look more, however now we have some volunteers offering to help.
That said, we already use the ee4j project parent POM, and use the nexus-staging-maven-plugin, so that piece at least is in place.

In addition, another concern looking ahead is how to fit the TCK execution into the pipeline. I'm happy to start without worrying about the TCK at all, for now... I would like to look to converting the TCK to Arquillian format later in the year. My only question is does that approach prevent us from leveraging any other EE4J common function other projects are using? Are other pipelines tightly coupled to TCK execution?

I realize there are some materials out there, e.g: https://www.youtube.com/watch?v=J5qLpTcRgH0
but anyway thought I'd ask for direction before diving in ourselves.

Thank you,
------------------------------------------------------
Scott Kurz
WebSphere Batch and Developer Experience
skurz@xxxxxxxxxx
--------------------------------------------------------

_______________________________________________
jakartaee-spec-project-leads mailing list
jakartaee-spec-project-leads@xxxxxxxxxxx
To unsubscribe from this list, visit https://urldefense.com/v3/__https://www.eclipse.org/mailman/listinfo/jakartaee-spec-project-leads__;!!GqivPVa7Brio!Nuee7rNIrHrbz_8Ah-ISNh9ObwG7GUDDnRLBcI5s8JV4y7d5e0n1OKnA4kjjik_NoQ$