Operations
Guides for operating a Sentinel deployment after it is installed.
In this section
- Monitoring & metrics: Health checks, Prometheus metrics, OpenTelemetry tracing, and log levels for Sentinel.
- Request logs & analytics: What Sentinel records in request logs, how long it keeps them, and how to search logs and read statistics.
- ClickHouse: Store Sentinel request logs in ClickHouse for high-volume deployments.
- Alerts: Configure Sentinel alert rules and deliver alerts by webhook, email, or web push.
- Audit logs: What Sentinel records in its audit log of administrative actions, how long it keeps entries, and how to search them.
- Snapshots & backups: Create, schedule, encrypt, and restore Sentinel database snapshots, and back up volumes as an alternative.
- Clustering & high availability: Run multiple Sentinel replicas: discovery, instance slots, shared secrets, and behavior during database or KV outages.
- Doctor & troubleshooting: Diagnose Sentinel with the doctor command and endpoint, and fix common problems by symptom.
- Performance & sizing: Sentinel v2 benchmark results and guidance for sizing CPU, memory, Redis, and ClickHouse.
- Schema migrations: How Sentinel creates and upgrades its database and ClickHouse tables, and how to apply migrations by hand with auto_migrate = false and schema sql.
- Updating Sentinel: Update Sentinel safely: version tags, database migrations on startup, rolling updates, and the release feed.