Jira Deployment Pipelines should be "grouped"
Everyone can contribute. Help move this issue forward while earning points, leveling up and collecting rewards.
This issue was created from this Slack thread (internal, good for 90 days).
Problem
Successive deployments are not appearing correctly in the Jira Development panel. Noticed first in #300031 (comment 1433009083), an earlier deployment failure will permanently display as a failure in the "Releases" section even after a later successful deploy.
| Releases section of the Jira issue shows failure | When viewing all deployments, the most recent deployment was successful |
|---|---|
 |
 |
Atlassian have clarified this is because the associated pipeline that we send has an ID that is unique per our modelling (it comes from PostgreSQL's id column):
The Jira docs for the endpoint mention:
Deployments are identified by the combination of pipelineId, environmentId and deploymentSequenceNumber, and existing deployment data for the same deployment will be replaced if it exists and the updateSequenceNumber of existing data is less than the incoming data.
So a combination of pipelineId and environmentId is what makes a deployment considered unique by Jira.
The intention for the Releases section to be updated to switch to showing a successful deployment after an earlier failure would be that the pipelineId was unchanged from the failure. Atlassian explained this as:
For example, if I deploy a service to staging then production, those are two separate deployments with the same pipelineId, different environmentId (one for stg, one for prod), and sequential deploymentSequenceNumbers.
This would then have the later successful deployment be identified as the same deployment that earlier failed, and their logic would understand this deployment to now have succeeded, and no longer be highlighting the earlier failure.
Proposal
The id of the pipeline we associate with a deployment will need to be a string that remains the same between successive deployments.
We'll need to determine an appropriate string to use here.
Note: We need to understand whether the original use of the PipelineEntity introduced a couple of years prior to deployments !49757 (merged) should remain unchanged (and we create a new pipeline entity to associate with deployments) or whether it also should change.
We must let Atlassian know once this issue is complete #431651 (comment 1646769026) as they have put a limit on the number of issue keys deployments are associated with for our app #431651 (comment 1646767502).