Metrics Reference

Dockward exposes Prometheus metrics at GET /metrics in Prometheus text exposition format. The endpoint is implemented without external dependencies — the format is hand-written to the stdlib net/http response writer.

See API Reference for endpoint details.

Metrics

Metric	Type	Labels	Description
watcher_updates_total	counter	service	Successful image deploys
watcher_rollbacks_total	counter	service	Rollbacks triggered after a failed deploy
watcher_restarts_total	counter	service	Auto-heal container restarts
watcher_failures_total	counter	service	Critical failures — restart exceeded max retries or rollback failed
watcher_service_healthy	gauge	service	1 if the service is healthy, 0 if not
watcher_service_blocked	gauge	service	1 if the service digest is blocked, 0 if not
watcher_poll_count_total	counter	—	Total poll cycles executed across all services
watcher_last_poll_timestamp_seconds	gauge	—	Unix timestamp of the most recent poll cycle
watcher_uptime_seconds	gauge	—	Seconds elapsed since dockward started

Example Output

# HELP watcher_updates_total Successful image updates
# TYPE watcher_updates_total counter
watcher_updates_total{service="myapp"} 4

# HELP watcher_rollbacks_total Rollbacks after failed deploy
# TYPE watcher_rollbacks_total counter
watcher_rollbacks_total{service="myapp"} 1

# HELP watcher_restarts_total Auto-heal restarts
# TYPE watcher_restarts_total counter
watcher_restarts_total{service="myapp"} 2

# HELP watcher_failures_total Critical failures
# TYPE watcher_failures_total counter
watcher_failures_total{service="myapp"} 0

# HELP watcher_service_healthy Service health state
# TYPE watcher_service_healthy gauge
watcher_service_healthy{service="myapp"} 1

# HELP watcher_service_blocked Digest blocked after rollback
# TYPE watcher_service_blocked gauge
watcher_service_blocked{service="myapp"} 0

# HELP watcher_poll_count_total Total poll cycles
# TYPE watcher_poll_count_total counter
watcher_poll_count_total 48

# HELP watcher_last_poll_timestamp_seconds Unix timestamp of last poll
# TYPE watcher_last_poll_timestamp_seconds gauge
watcher_last_poll_timestamp_seconds 1.740567321e+09

# HELP watcher_uptime_seconds Seconds since start
# TYPE watcher_uptime_seconds gauge
watcher_uptime_seconds 14412

Scrape Configuration

Add the following to your Prometheus scrape_configs:

scrape_configs:
  - job_name: dockward
    static_configs:
      - targets: ["localhost:9090"]