Metrics

Harbor Sync Controller exposes prometheus metrics. You configure the listen address / port via -metrics-addr. The following metrics are available:

metric name	type	labels	description
`http_request_duration_seconds`	histogram	`code,method,path`	keeps track of the duration API requests towards harbor
`harbor_matching_projects`	gauge	`config,selector_type,selector_project_name`	total number of matching projects per HarborSyncConfig
`harbor_robot_account_expiry`	gauge	`project,robot`	the date after which the robot account expires, expressed as Unix Epoch Time
`harbor_sync_sent_webhooks`	gauge	`config,target,status_code`	The number of webhooks sent

Alerts

Here are example alerts

groups:
- name: harbor_rules
  rules:
  - alert: HarborAccountExpires
    expr: (harbor_robot_account_expiry - time()) / 86400 < 14
    for: 10m
    labels:
      severity: warning
    annotations:
      summary: "CRITICAL: harbor robot account '{{ $labels.robot }}' in project '{{ $labels.project }}' expires in less than 14d"
      description: "harbor robot account expires soon"
  - alert: HarborOutgoingWebhooksFailed
    expr: sum(increase(harbor_sync_sent_webhooks{status_code!="200"}[1h])) by (target) > 0
    labels:
      severity: critical
    annotations:
      summary: "CRITICAL: harbor outgoing webhook failed: '{{ $labels.target }}'"
      description: "harbor does not deliver the robot account information correctly"