General Notes

• Make sure to initially set-up your environment.

• Use the command help to get a list of all commands in the release tools.

• After fixing a problem use workspace cleanup to clean-up any mess left behind by the previous step.

The release process

The Pre-Release tasks and Post-Release tasks phases of making a release (opening tickets/closing tickets) is still done manually.

The Build and publish the release phase is what’s covered by the CI server.

🍃 Pre-Release tasks

1. Ensure all work on CVEs potentially contained in the release is done (incl. backports etc.).

2. Upgrade dependencies in Spring Data Build parent pom (mind minor/major version rules).

3. All release tickets are present (CLI: $ tracker releasetickets $trainIteration)

4. Review open tickets for release.

5. Self-assign release tickets (CLI: $ tracker prepare $trainIteration)

6. Announce release preparations to mailing list (https://groups.google.com/forum/#!forum/spring-data-dev)

7. 🚥 Continue with Build and publish the release

🌿 Build and publish the release

Action	Command
Check out this repository from GitHub	$ git clone [email protected]:spring-projects/spring-data-release.git
Checkout the main branch	$ git checkout main
Edit ci/release.properties and assign the version number of the release train you wish to release to release.version. (Make sure it’s NOT commented out!)
Commit the change to this repository	$ git add ci/release.properties && git commit
Switch to the release branch	$ git checkout release
Set the HEAD to main to point the content to what is on main	$ git reset --hard main
🚨 To trigger 🛣 the release git push the release branch	$ git push origin release
Go to https://jenkins.spring.io/view/SpringData/job/spring-data-release-release/, Navigate to the release branch. A job should have started. Click on the active job, and then click on Open Blue Ocean.	Monitor the release process.

After preparing and building (approx. 20 minutes), the Jenkins pipeline will pause, indicating it’s time to smoke test the staged release, and put up a prompt: Proceed or Abort.

🚨 If you click on Proceed, then it will continue on to the release conclude/github push/release distribute phase.

• For a Maven central release, go to https://s01.oss.sonatype.org/ and login.

  • After logging in, you will find a closed Spring Data Release repository.

  • Click the repository and then choose "Release".

• For an artifactory release, the release will already have been staged and promoted, so there is nothing more to do.

• 🚥 Continue with Post-release tasks

Background

The first CI phase includes: release prepare and release build and may take twenty minutes or so.

• For a maven central release, you can visit https://s01.oss.sonatype.org/, login, and inspect the staged Spring Data Release. All the Spring Data modules will inside one repository.

• For a milestone/RC release, you can visit https://repo.spring.io/. All the Spring Data modules will be there, each module a separate "build".

At this point in time, we don’t have official Spring Data Release smoke testing configured, so you can simply click Proceed if there were no issues.
For milestone/RC releases, Spring Data Release tools automatically promotes them in Artifactory, so if you have to delete a build, you’ll ALSO have to go and delete each module individually.

🍀 Post-Release tasks

Action	Command
Finally, push the commit that changed release.properties to main	$ git push origin main (where you’ve checked out spring-data-release)
Close release tickets.	$ tracker close $trainIteration
Create new release versions and tickets for upcoming version	$ tracker setup-next $trainIteration.next
Update versions in Sagan. $targets is given as comma separated lists of code names, without spaces. E.g. Moore,Neumann	$ sagan update $releasetrains
Create list of docs for release announcements	$ announcement $trainIteration
Announce release (Blog, Twitter) and notify downstream dependency projects as needed.	N.A.

Congratulations 🥳 You completed the release ❤️.

Appendix

One Time Setup

Infrastructure requirements

• Ensure you have the credentials for buildmaster accounts on https://repo.spring.io.

• Ensure you have the credentials for https://oss.sonatype.org (to deploy and promote GA and service releases, need deployment permissions for org.springframework.data) in settings.xml for server with id sonatype.

Both are available in the Spring/Pivotal Last Pass repository.

Prepare local configuration and credentials

Add an application-local.properties to the project root and add the following properties:

• git.username - Your GitHub username.

• git.password - Your GitHub Password (or API key with scopes: public_repo, read:org, repo:status, repo_deployment, user when using 2FA).

• git.author - Your full name (used for preparing commits).

• git.email - Your email (used for preparing commits).

• maven.mavenHome - Pointing to the location of your Maven installation.

• deployment.username - Your Artifactory user.

• deployment.api-key - The Artifactory API key to use for artifact promotion.

• deployment.password - The encrypted Artifactory password..

• gpg.keyname - The GPG key name.

• gpg.passphrase - The password of your GPG key.

• gpg.executable - Path to your GPG executable, typically /usr/local/MacGPG2/bin/gpg2 or /usr/local/bin/gpg.

• sagan.key - Sagan authentication token. Must be a valid GitHub token. Can be the same as git.password when using a GitHub token as password.

After that, run the verify command ($ verify) to verify your settings (authentication, correct Maven, Java, and GPG setup).

See application-local.template for details.

Detailed commands performed by spring-data-release-cli

Action	Command
All release tickets are present	$ tracker releasetickets $trainIteration
Self-assign release tickets	$ tracker prepare $trainIteration
Release the binaries
	$ release prepare $trainIteration
Build the artefacts and push them to the appropriate maven repository	$ release build $trainIteration
	$ release conclude $trainIteration
Push the created commits to GitHub	$ github push $trainIteration
Push new maintenance branches if the release version was a GA release (X.Y.0 version)	$ git push $trainIteration.next
Distribute documentation and static resources from tag	$ release distribute $trainIteration
Post-release tasks
Close JIRA tickets and GitHub release tickets.	$ tracker close $trainIteration
Create new release versions and tickets for upcoming version	$ tracker setup-next $trainIteration.next
Update versions in Sagan. $targets is given as comma separated lists of code names, without spaces. E.g. Moore,Neumann	$ sagan update $releasetrains
Create list of docs for release announcements	$ announcement $trainIteration

Utilities

GitHub Labels

ProjectLabelConfiguration contains a per-project configuration which labels should be present in a project. To apply that configuration (create or update), use:

$ github update labels $project

Dependency Upgrade

ProjectDependencies contains a per-project configuration of dependencies.

Workflow:

• Check for dependency upgrades $ dependency check $trainIteration

Reports upgradable dependencies for Build and Modules and creates dependency-upgrade-build.properties file. Edit dependency-upgrade-build.properties to specify the dependency version to upgrade. Removing a line will omit that dependency upgrade.

• Apply dependency upgrade with $ dependency upgrade $trainIteration. Applies dependency upgrades currently only to Spring Data Build.

• Report store-specific dependencies to Spring Boot’s current upgrade ticket (sample) $ dependency report $trainIteration

CI Properties Distribution

To distribute ci/pipeline.properties across all modules use:

$ infra distribute ci-properties $trainIteration