Configure Metrics

Pomerium Enterprise uses Prometheus as a metrics collection back-end. You can configure Pomerium and the Console to talk to an existing Prometheus server, or configure the embedded Prometheus backend.

TIP

For production deployments, we suggest using a dedicated Prometheus instance.

Prepare Pomerium

1. In the Pomerium config.yaml, define the metrics_address key to a network interface and/or port. For example:

   metrics_address: 192.0.2.31:9999
   

   The example above has Pomerium providing metrics at port 9999 on an IP address reachable by the Pomerium Console service.

   If you're running Pomerium Enterprise in a distributed environment where the IP address is not known at the time of deployment, you can use the resolvable FQDN of the Pomerium host (pomerium0.internal.mycompany.com, for example), or override this key with the environment variable METRICS_ADDRESS. We do not recommend exposing this endpoint to public traffic as it can contain potentially sensitive information.

External Prometheus

1. Add the listener to your Prometheus configuration, usually via prometheus.yml:

   - job_name: 'Pomerium'
      scrape_interval: 30s
      scrape_timeout: 5s
      static_configs:
         - targets: ['192.0.2.10:9999']
   
   

2. Reload (opens new window) the Prometheus configuration:

   curl -i -XPOST path.to.prometheus:port/-/reload
   

3. In the Pomerium Enterprise config.yaml file, define the prometheus_url key to point to your Prometheus instance(s):

   prometheus_url: http://192.168.122.50:9090
   

4. Restart the Pomerium and Pomerium Enterprise services. You should now see route traffic data in the Enterprise Console:

   

Embedded Prometheus

To take advantage of Prometheus embedded in Pomerium Enterprise, edit /etc/pomerium-console/config.yaml:

prometheus_data_dir: /var/lib/pomerium-console/tsdb

The directory path can be any location that the pomerium system user can write to. The example above uses the default location created by the OS packages.