Metrics and logs

Metrics

The service's Metrics tab shows:

• CPU usage, against the service's vCPU limit;
• Memory usage, against its memory limit;
• Network traffic;
• Replicas running.

Choose a range of 1 hour, 6 hours, 24 hours or 7 days. Live switches to a view that updates every 5 seconds, useful while you load-test or watch a deployment.

Because Fjordrun bills actual usage, the CPU and memory graphs are also what you pay for. The service's usage line shows the cost so far in the current billing period.

Logs

Log	Contains	Where
Build logs	Clone, the builder's detection, install and build output, image push	The deployment
Deploy logs	Rollout of the new version	The deployment
Runtime logs	Your app's stdout and stderr	The service, and the project's Observability tab

Build and deploy logs stream live while a deployment runs. Runtime logs can be filtered by text and by level (Info, Warn, Error); the level is taken from the log line.

Write logs to stdout and stderr, not to files. One line per event (or JSON lines) makes filtering easiest.

CLI

fjordrun logs           # recent runtime logs of the linked service
fjordrun logs -f        # follow
fjordrun logs --build   # build logs of the latest deployment

API

• GET /api/v1/services/{id}/logs?search=timeout&level=error&limit=200: runtime logs.
• GET /api/v1/deployments/{id}/logs: build and deploy logs of a deployment; .../logs/stream streams them as server-sent events.
• GET /api/v1/services/{id}/metrics?range=24h and .../metrics/live (server-sent events).

See API tokens for authentication.

Activity

Who deployed, changed variables, added domains or changed billing is recorded in the project and team activity feeds.