Alert Visibility
Overview
OpsWorker captures and normalizes every alert that arrives from your connected monitoring systems, providing a unified view of all alert activity across clusters and namespaces.
All alerts are recorded, whether or not they trigger an investigation.
Supported Alert Sources
OpsWorker accepts alerts from:
- Prometheus AlertManager
- Datadog
Grafana Alerting is supported because it sends alerts in AlertManager format; it is not a separate native integration. Datadog is an alert source only (OpsWorker does not query Datadog).
Unified Alert View
The alert timeline in the OpsWorker portal shows:
- All alerts from the supported sources in one place
- Chronological timeline of alert activity
- Alert details including name, severity, namespace, labels, and source
Filtering and Search
Filter alerts to find what you need:
| Filter | Options |
|---|---|
| Severity | By severity level |
| Namespace | Any namespace in your cluster |
| Cluster | Filter by specific cluster |
| Time range | Custom date/time selection |
| Search | Text search across alert names and labels |
Dashboard
The portal Dashboard reports a focused set of metrics:
- Alerts per cluster per day, and overall counts by status
- Finished investigations per cluster per day
- Time saved in minutes
The Dashboard does not aggregate MTTR, accuracy, ROI, or severity-by-source breakdowns. Investigation feedback is collected only as thumbs up or down in Slack and is not aggregated into the portal.
Why Visibility Matters
Even alerts that don't trigger investigations provide value:
- Identify noisy alert sources that could be tuned or filtered with alert rules
- Audit alert coverage: see which namespaces and services are sending alerts
- Inform alert rule creation: identify alerts that should trigger auto-investigation
Next Steps
- Noise Reduction: Filter and deduplicate alerts
- Daily Digest: Automated daily per-cluster summaries