Healthchecks

Both the hub and agent have health commands that can be used for Docker healthchecks.

They add a bit of CPU overhead, so we recommend using an interval of 60s or more.

Agent example

The agent healthcheck verifies the agent is running properly, not necessarily that it's connected to the hub.

services:
  beszel-agent:
    healthcheck:
      test: ['CMD', '/agent', 'health']
      interval: 120s

Hub example

The hub healthcheck tests if the /api/health endpoint returns a 200 status code.

services:
  beszel-hub:
    healthcheck:
      # The URL is relative to the container, not the host
      test: ['CMD', '/beszel', 'health', '--url', 'http://localhost:8090']
      start_period: 5s # Check 5 seconds after the container starts
      interval: 120s # Then check every 120 seconds after that