Prometheus

Autometrics is built on top of Prometheus (opens in a new tab) - a popular open source timeseries database and alerting system.

To see the metrics your instrumented app generates you will need to deploy a Prometheus instance alongside it. Prometheus is pull-based, meaning that once deployed it will poll your app on an endpoint you specify to sample the metrics.

Prometheus locally

To get started and understand how Prometheus works you can run it locally:

Run Prometheus using Autometrics CLI (Recommended)Run Prometheus manually

Prometheus in production

If you've already tried out Autometrics with Prometheus locally and want to set it up in production, see instructions below for specific deploy targets.

Deploying Prometheus with Docker Deploying Prometheus in Kubernetes Deploying Prometheus on Northflank Deploying Prometheus on Railway Configuring Prometheus on Fly.io

Autometrics Stack Docker