Instant Metrics

In traditional setups, monitoring specific processes usually requires installing a sidecar agent like process-exporter or writing custom scripts to parse ps output.

Super eliminates this friction by providing a built-in Prometheus metrics endpoint.

Zero Configuration

As soon as superd starts, metrics are available. No plugins, no configuration needed.

When API auth is off (default loopback OSS without auth_secret):

$ curl http://localhost:9002/metrics

When auth is on (auth_secret or the security plugin), scrape with Bearer:

$ curl -H "Authorization: Bearer $SUPER_TOKEN" http://localhost:9002/metrics

Sample Output:

# HELP super_process_up Process status
# TYPE super_process_up gauge
super_process_up{id="...",name="api-server",group="backend"} 1
super_process_cpu_percent{id="...",name="api-server",group="backend"} 2.5
super_process_memory_bytes{id="...",name="api-server",group="backend"} 45000000
super_process_uptime_seconds{id="...",name="api-server",group="backend"} 1200
super_process_restart_count{id="...",name="api-server",group="backend"} 0

Integrating with Prometheus

Add Super to your prometheus.yml scrape configs.

Without auth (loopback OSS, no auth_secret):

scrape_configs:
  - job_name: 'super'
    static_configs:
      - targets: ['127.0.0.1:9002']

With auth enabled, pass the same Bearer secret / Access Token:

scrape_configs:
  - job_name: 'super'
    authorization:
      type: Bearer
      credentials: 'your-auth-secret-or-sk-token'
    static_configs:
      - targets: ['127.0.0.1:9002']

Useful PromQL Queries

With these metrics, you can create a powerful Grafana dashboard in minutes.

1. Detect Flapping Services

Find services that have restarted more than 5 times in the last hour.

increase(super_process_restart_count[1h]) > 5

2. High Memory Usage Alert

Alert if any process uses more than 1GB of RAM.

super_process_memory_bytes > 1073741824

3. Service Down Alert

Check if any managed service is not running.

super_process_up == 0

Plugin metrics 💎

With licensed plugins loaded, superd exports additional metrics (e.g. from notify):

• super_cgroup_enforced_total: Number of processes with active resource limits.
• super_notify_sent_total: Statistics on successful/failed webhook notifications.