wouldd opened a new issue, #7193: URL: https://github.com/apache/incubator-devlake/issues/7193
### Search before asking - [X] I had searched in the [issues](https://github.com/apache/incubator-devlake/issues?q=is%3Aissue) and found no similar issues. ### What happened I assume that the issue I'm seeing is somehow unique to me using the azuredevops_go plugin and the webhook for reporting deployments from octopus. However the 'bug' I've identified looks to me like it would effect everyone. So probably the real issue is somewhere else but all I can say is that for me the DORA metrics are all wildly out because it relates every commit that has ever been made against a repo to the most recent deploy. This means the cycle time calculation for each commit is showing as just the age of that commit until the most recent deploy. I believe the offending line is this one: https://github.com/apache/incubator-devlake/blob/cb69f35b5319f57dd3d6c8534121a6ca069cd175/backend/plugins/dora/tasks/change_lead_time_calculator.go#L215C2-L215C3 the commits_diffs table has 3 columns, commit_sha, old_commit_sha and new_commit_sha. For me I need the filter in this query to filter where the mergeSha = cd.new_commit_sha in order to work correctly. I've tested this in my setup and that change has the expected behaviour. However, I would obviously expect this to be broken for everyone doing DORA metrics at all, and it seems unlikely no one else noticed this. So I guess maybe the intention of that commits_diffs table is to have information in it differently and perhaps the azure plugin is populating it incorrectly? ### What do you expect to happen In my case I have lots of quite old 4+years minimum repos. but I've only just started feeding deploy events in. so I would expect only to have data for the relatively few new commits since this went into place. and I would expect the cycle time calculations for those to be right. ### How to reproduce configure a long standing repo with lots of historical commtis against it. create a webhook call for the most recent commit/commtis in teh repo and expect to see dora calculations for those commits only. However you should see that it links all of the old commits to the new deployment. Indeed when I run with a version in which I changed the above filter check then I get the behavior I expect. all the numbers make sense and the pr_cycle_time calculations show up numbers for only commits that have been reported as part of deployments ### Anything else Happens every time for all commits. ### Version master ### Are you willing to submit PR? - [X] Yes I am willing to submit a PR! ### Code of Conduct - [X] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
