Installation
This page installs the Momento Valkey Operator from the v0.6.0 GitHub release artifacts and verifies the rollout. It assumes the cluster-admin access described in Prerequisites.
1. Apply the CRDs
kubectl apply -f https://github.com/momentohq/valkey-operator/releases/download/v0.6.0/crds.json
This registers all five custom resource definitions: ValkeyImage, ValkeyConfig, ValkeyRole, ValkeyCluster, and ValkeyNode.
2. Deploy the operator
kubectl apply -f https://github.com/momentohq/valkey-operator/releases/download/v0.6.0/operator.yaml
The operator image is pulled from Docker Hub at gomomento/valkey-operator.
What this installs
operator.yaml creates:
| Object | Name | Notes |
|---|---|---|
| Namespace | valkey-operator | Holds every object below. |
| ServiceAccount | valkey-operator | Identity the operator Deployment runs as. |
| ClusterRole + ClusterRoleBinding | valkey-operator | Grants the operator the permissions it needs across all namespaces. See RBAC. |
| ConfigMap | valkey-operator-config | Optional process settings, only defaultZoneSpread. See Operator configuration. |
| Deployment | valkey-operator | Runs the operator container. |
The Deployment runs a single replica with no leader election; one replica is the entire concurrency control, by design. If that pod restarts or is briefly unavailable during an upgrade, reconciliation pauses, but no running Valkey cluster is affected. Valkey clusters serve traffic independently of the operator process. See Architecture for why the stateless, single-replica design holds up in practice.
The operator container is a distroless image (no shell or package manager) running a single process. The manifest sets no resource requests or limits and no probes on the Deployment. If your platform requires them (for capacity planning, quota, or scheduling policy), add them to the Deployment after installing. The operator's footprint is that of a single lightweight control loop, and the docs make no specific sizing claim for it. For what to monitor on the operator itself, see Monitoring.
Verify the rollout
kubectl -n valkey-operator rollout status deployment/valkey-operator
This returns once the operator pod is Running and ready.
Confirm the CRDs registered:
kubectl get crd | grep valkey.gomomento.com
You should see five entries: valkeyimages, valkeyconfigs, valkeyroles, valkeyclusters, and valkeynodes, all under valkey.gomomento.com.
Next steps
- Continue to the Quickstart to provision your first Valkey cluster.
- Review Operator configuration if you need to change
defaultZoneSpreadbefore teams start creating clusters. - Bookmark Operator upgrades for when the next release ships.
- Bookmark Uninstall for the reverse procedure and the order it must run in.