Properly Setting up Devlake for Measuring DORA Metrics in a Multi-Repo Setup

[Audefleda]

Hey there,

I have the following setup:

Our software is build in a complex setup up multiple repos and pipelines in Azure Devops. The "final" pipeline is creating an installer package that is delivered to an internal platform for download. A new version of this installer package is what constitutes a "production deployment" in our context. All previous pipelines are just creating binaries, that are then put together into the installer package in the final pipeline.

I'm using the build in plugin for connecting the repos, but are using webhooks for providing Incidents and Deployments.

My question is now: Which of the repos and deployment pipeline do I need to connect to Apache Devlake? My initial idea was to only connect the final installer pipeline and repo. But would that be sufficient to properly calculate the Lead Time for Changes?

On the other hand, If I connect all the previous repos and pipelines, how can I provide the necessary information for DevLake about which pull request ended in which pipeline run of the installer?

Thanks a lot for your support!

[dosubot[bot]]

Hey @Audefleda👋, I'm here to help you with any bugs, questions, or contributions you have. Let's tackle this together!

To properly set up Apache DevLake for measuring DORA metrics in your multi-repo setup, you should consider the following:

1. Connect the Final Pipeline and Repo: Since the final pipeline creates the installer package that constitutes a "production deployment," connecting this pipeline and its associated repo is crucial. This will allow you to track the deployments accurately.

2. Connect All Repos and Pipelines: To calculate the Lead Time for Changes accurately, you should connect all the repos and pipelines. This ensures that DevLake can track the entire lifecycle of changes from code commit to deployment.

3. Link Pull Requests to Pipeline Runs: Use webhooks or direct connections to provide information about which pull request ended in which pipeline run. This can be done by configuring your CI/CD system to send data to DevLake, indicating the association between pull requests and specific pipeline runs.

By connecting all relevant repos and pipelines and ensuring that pull requests are linked to pipeline runs, you can accurately measure DORA metrics, including Lead Time for Changes, in your setup.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}