12-Factor Web development: Support for custom Prometheus scrape configurations

announcements and community
1

With the latest release of paas-charm, we are improving the Runtime Configuration support.

This feature allows you to extend certain capabilities of your charm without modifying the Python code, simply by dropping a paas-config.yaml file alongside your charmcraft.yaml.

The first major capability enabled by this configuration is Prometheus custom metrics collection.

Does your application run a service that exposes metrics on a different port? Do you have a Celery Beat scheduler exporter that needs a specific scrape target? You can now define these explicitly.

The following snippet is an example of how to use the new paas-config.yaml to configure Prometheus targets:

prometheus:
  scrape_configs:
    # Application metrics from all units
    - job_name: "flask-app-custom"
      metrics_path: "/metrics"
      static_configs:
        - targets:
            - "*:8081"
          labels:
            app: "flask"
            env: "example"

    # Metrics scraped from only the scheduler unit
    - job_name: "flask-scheduler-metrics"
      metrics_path: "/metrics"
      static_configs:
        - targets:
            - "@scheduler:8082"
          labels:
            role: "scheduler"

For the full documentation, see:

2 Likes