Monitoring

Eyrie monitors let you track the health of your deployments with automated checks and threshold-based alerts.

Monitor Types

• http_request — sends an HTTP request to the deployment’s URL and checks for a successful response.

• Resource-based — checks CPU, memory, disk usage, SSL certificate expiration, or time since last backup.

Creating a Monitor

1. Open a deployment at /_eyrie/p/<slug>/dg/<group>/d/<deploy>.

2. Navigate to the Monitors tab.

3. Click Add Monitor and configure:

   • Type — http_request or a resource metric.

   • Thresholds — warning and error levels (e.g. CPU > 80% = warning, CPU > 95% = error).

Scopes with monitor = True automatically create default monitors when a deployment is deployed.

Monitor States

Each monitor reports one of four states:

• OK — all checks pass.

• Warning — a threshold has been crossed but the service is operational.

• Error — a critical threshold has been crossed or the check failed.

• Unknown — the monitor has not run yet or produced an inconclusive result.

The deployment card shows a rolled-up health indicator based on the worst monitor state.

Thresholds

Available threshold metrics:

• CPU % — container CPU utilization.

• Memory % — container memory utilization.

• Disk % — persistent volume usage.

• SSL Expiration — days until the TLS certificate expires.

• Last Backup Hours — hours since the most recent successful backup.

Running a Monitor Manually

Click the Run Now button on any monitor to execute an immediate check outside the normal schedule.