How to set up Sleuth and deliver DORA metrics in minutes

October 9, 2021

Automatically track DORA metrics in about a minute? Let's do it. Start by visiting Sleuth at sleuth.io, and clicking on Try Now for Free. Type in your organization name, and then sign up with Google, GitHub, or another provider. I'm going to choose GitHub because that's where my source code is. Sleuth will ask for permissions to mainly view code, but a few other things as well. This page will be different depending on what provider you chose. Next, Sleuth will ask you to name your project. I'm going to take the default. Next, let's add a code repository. Notice, Sleuth allows you to match branches to environments. So for example, if you have a production branch, you can map that to the production environment. I use master for everything.

Finally, Sleuth wants to know how to determine the health of your service. Sleuth supports all sorts of services, and even custom integrations. But we're going to choose Sentry. Sleuth can track the number of errors reported this Sentry, and then, through anomaly detection, determine if the service is healthy or not. And that's it. Sleuth will now track closed-pull requests as deployments and report the door metrics from there. However, I highly recommend this next step if you want the most accurate metrics. Go to the code deployment, and configure a webhook to tell Sleuth exactly when you do a deployment. In my project, if the file looks like this, we use Fabric to deploy, and so we put this curl statement to tell Sleuth when we deploy what code to what environment. For an overview of metrics Sleuth tracks and how, see this video. For a deep dive in how you can measure and reduce change lead time, see this video.