Skip to main content

When to use this

Use Sentinel when the deployment should support failover from one FalkorDB replica to another. The Admin UI maps this flow to:
  • spec.topology: replication
  • spec.replicas from the FalkorDB replica count
  • spec.sentinel.enabled: true
  • spec.sentinel.replicas from the Sentinel node count

Create from Admin UI

  1. Open Deployments > Create Deployment.
  2. Choose Standalone.
  3. Choose QA or Prod to start from a Sentinel-enabled template, or choose Full Setup and enable Sentinel manually.
  4. In FalkorDB, set the FalkorDB replica count. Prod requires at least 2.
  5. In Sentinel, enable Sentinel and set an odd number of Sentinel replicas. The UI defaults to 3.
  6. Configure Sentinel CPU, memory, storage, and optional node affinity.
  7. Optionally add a Sentinel service in Networking, or set an external domain for stable per-pod DNS names outside the cluster — see Networking.
  8. Complete auth, TLS, backup, parameters, and review.

Template defaults

Failover status

The current Pods tab displays a Switchover confirmation for the primary pod, but this UI path currently only shows a success toast. There is no backend switchover operation wired in this codebase yet. Until that is implemented, handle manual failover through your Kubernetes/KubeBlocks operating procedure and verify the resulting pod roles in the Admin UI.

Validate

In the Admin UI, verify Pods, Operations, Logs, and Metrics.