Provides names for our pipelines
What does this MR do?
- For all pipelines that do not match the rule, they will not be named - effectively no change from prior behavior
- If we detect this is a trigger from release-tools, this must be an
Auto-Deploy pipeline, the name should look something like:
Deployment QA pipeline - 15.8.202301061520-1fe404583c3.5ebc0742181
- Note that I am not including the environment nor stage in the name - this will be sufficed by the name of the project where the pipeline is run
-
$DEPLOY_VERSION
comes fromrelease-tools
as one of the variables sent to this project upon pipeline trigger
- If this is a scheduled pipeline, let's name it. This duplicates the data as the pipeline is labeled, but the commit name is not a good representation of WHY the pipeline runs.
- The variable
$QA_PIPELINE_NAME
was chosen as this is relatively unique which would be useful if this is passed in from other projects- We are attempting to do this elsewhere, such as from
k8s-workloads/gitlab-com
where we trigger QA from - Important docs on variable inheritance: https://docs.gitlab.com/ee/ci/variables/#cicd-variable-precedence
- We are attempting to do this elsewhere, such as from
- We then include the default ruleset for workflow rules to ensure that
all other pipelines styles are indeed created
- A rule is added to enable Engineers to create a pipeline even when an MR is opened
Changelog: added
Addresses: gitlab-com/gl-infra/delivery#2754 (closed)
Check-list
-
Verify the change in all affected pipelines.
Edited by John Skarbek