Bitbucket Cloud auto-versioning post-processing

You can delegate the post-processing to a Bitbucket pipeline.

Yontrack triggers a custom: pipeline with the variables describing the upgrade, reports the link to the pipeline run in the auto-versioning audit, and waits for the pipeline to complete. The post-processing fails when the pipeline does not complete successfully, or does not complete in time.

Note

The Bitbucket Cloud configuration used to trigger the pipeline needs a token with the pipelines write scope. See the Bitbucket Cloud configuration.

Settings

There is a global configuration, and there is a specific configuration at the branch level (in the postProcessingConfig parameter).

For the global configuration, you can go to Settings > Bitbucket Cloud Auto Versioning Post Processing and define the following attributes:

• Configuration - Default Bitbucket Cloud configuration to use for the connection
• Workspace - Default workspace of the repository containing the pipeline
• Repository - Default repository containing the pipeline
• Pipeline - Name of the custom pipeline containing the post-processing (like yontrack-auto-versioning)
• Branch - Branch to run the pipeline on, main by default
• Retries - The amount of times we check for the completion of the post-processing pipeline, 10 by default
• Retry interval - The time (in seconds) between two checks for the completion of the post-processing pipeline, 30 by default and never less than 10 seconds

Yontrack waits at most Retries x Retry interval for the pipeline to complete: 5 minutes by default.

The settings can also be set as code:

ontrack:
  config:
    settings:
      bitbucket-cloud-av-post-processing:
        config: bitbucket-cloud
        workspace: my-workspace
        repository: post-processing
        pipeline: yontrack-auto-versioning
        branch: main
        retries: 10
        retriesDelaySeconds: 30

Branch configuration

The postProcessingConfig property at the branch level accepts the following parameters:

Parameter	Default value	Description
dockerImage	Empty	This image defines the environment for the upgrade command to run in
dockerCommand	Empty	Command to run in the Docker container
commitMessage	Empty	Commit message to use to commit and push the result of the post-processing
config	Settings	Bitbucket Cloud configuration to use for the connection
workspace	Settings	Workspace of the repository containing the pipeline
repository	Settings	Repository containing the pipeline
pipeline	Settings	Name of the custom pipeline to run
branch	Settings	Branch to run the pipeline on

The config, workspace, repository and pipeline must be defined, either at the branch level or in the settings, otherwise the post-processing fails before any pipeline is triggered.

The dockerImage, dockerCommand and commitMessage parameters, and the branch, are templated using the auto-versioning templating context. For example:

commitMessage: "Upgrading to ${VERSION}"

Example of a simple configuration relying on the global settings:

postProcessing: bitbucket-cloud
postProcessingConfig:
  dockerImage: eclipse-temurin:21
  dockerCommand: ./gradlew dependencies --write-locks
  commitMessage: "Resolving the dependency locks"

Example of a configuration running a pipeline defined in the target repository itself:

postProcessing: bitbucket-cloud
postProcessingConfig:
  dockerImage: eclipse-temurin:21
  dockerCommand: ./gradlew dependencies --write-locks
  commitMessage: "Resolving the dependency locks"
  workspace: my-workspace
  repository: my-repository
  pipeline: yontrack-auto-versioning

Pipeline variables

Yontrack triggers the pipeline with the following variables:

Variable	Description
REPOSITORY	URI of the repository to process
UPGRADE_BRANCH	Branch containing the changes to process
DOCKER_IMAGE	This image defines the environment for the upgrade command to run in
DOCKER_COMMAND	Command to run in the Docker container
COMMIT_MESSAGE	Commit message to use to commit and push the result of the post-processing
VERSION	The version which is upgraded to

The code below shows an example of a bitbucket-pipelines.yml custom pipeline suitable for post-processing:

pipelines:
  custom:
    yontrack-auto-versioning:
      - variables:
          - name: REPOSITORY
          - name: UPGRADE_BRANCH
          - name: DOCKER_IMAGE
          - name: DOCKER_COMMAND
          - name: COMMIT_MESSAGE
          - name: VERSION
      - step:
          name: Post-processing
          services:
            - docker
          script:
            # Credentials able to push to the repository to process, as secured repository variables
            - git clone --branch "${UPGRADE_BRANCH}" "https://x-token-auth:${PUSH_TOKEN}@${REPOSITORY#https://}" work
            - cd work
            - docker run --rm -v "$PWD:/work" -w /work "${DOCKER_IMAGE}" sh -c "${DOCKER_COMMAND}"
            - git config user.name "Yontrack post-processing"
            - git config user.email "yontrack@example.com"
            - git add --all
            - git commit -m "${COMMIT_MESSAGE} [skip ci]"
            - git push origin "HEAD:${UPGRADE_BRANCH}"

Important

• all the variables Yontrack sends must be declared in the custom pipeline
• committing and pushing the changed files to UPGRADE_BRANCH is required for the post-processing to be considered complete
• [skip ci] in the commit message prevents the push from starting the other pipelines of the repository

The rest of the pipeline can be adapted at will.