You are looking at the documentation of a prior release. To read the documentation of the latest release, please
visit here.
New to KubeDB? Please start here.
Restart HanaDB
This guide shows how to restart a HanaDB database using a HanaDBOpsRequest of type Restart. KubeDB
performs a rolling restart, one pod at a time. For a System Replication cluster the primary is
restarted last to minimize avoidable failovers.
Note: The YAML files used in this tutorial are stored in docs/examples/hanadb/restart folder in the GitHub repository kubedb/docs.
Before You Begin
- Install the KubeDB Provisioner and Ops-manager operators following the steps here.
- Create a namespace:
$ kubectl create ns demo
namespace/demo created
Deploy a HanaDB System Replication Cluster
apiVersion: kubedb.com/v1alpha2
kind: HanaDB
metadata:
name: hanadb-cluster
namespace: demo
spec:
version: "2.0.82"
replicas: 2
storageType: Durable
topology:
mode: SystemReplication
systemReplication:
replicationMode: fullsync
operationMode: logreplay_readaccess
podTemplate:
spec:
containers:
- name: hanadb
resources:
requests:
cpu: "1500m"
memory: "8Gi"
limits:
cpu: "4"
memory: "14Gi"
storage:
accessModes: ["ReadWriteOnce"]
resources:
requests:
storage: 64Gi
storageClassName: local-path
deletionPolicy: WipeOut
$ kubectl apply -f https://github.com/kubedb/docs/raw/v2026.7.10/docs/examples/hanadb/restart/system-replication-ops.yaml
hanadb.kubedb.com/hanadb-cluster created
Wait until hanadb-cluster is Ready. Note the current pod ages and which pod is primary:
$ kubectl get pods -n demo -l app.kubernetes.io/instance=hanadb-cluster -L kubedb.com/role
NAME READY STATUS RESTARTS AGE ROLE
hanadb-cluster-0 2/2 Running 0 14m primary
hanadb-cluster-1 2/2 Running 0 14m secondary
hanadb-cluster-arbiter-0 1/1 Running 0 8m arbiter
Create a Restart HanaDBOpsRequest
apiVersion: ops.kubedb.com/v1alpha1
kind: HanaDBOpsRequest
metadata:
name: hdbops-restart
namespace: demo
spec:
type: Restart
databaseRef:
name: hanadb-cluster
timeout: 30m
apply: Always
$ kubectl apply -f https://github.com/kubedb/docs/raw/v2026.7.10/docs/examples/hanadb/restart/restart.yaml
hanadbopsrequest.ops.kubedb.com/hdbops-restart created
Here spec.apply: Always lets the restart proceed even if the database is not Ready, which is useful
for recovering an unhealthy database.
Verify the Restart
$ kubectl get hdbops -n demo hdbops-restart
NAME TYPE STATUS AGE
hdbops-restart Restart Successful 7m32s
$ kubectl describe hdbops -n demo hdbops-restart
...
Status:
Conditions:
Message: HanaDBOpsRequest has started to restart HanaDB nodes
Reason: Restart
Status: True
Type: Restart
Message: Successfully paused database
Reason: DatabasePauseSucceeded
Status: True
Type: DatabasePauseSucceeded
Message: Successfully restarted HanaDB nodes
Reason: RestartNodes
Status: True
Type: RestartNodes
Message: Successfully completed restart for HanaDB.
Reason: Successful
Status: True
Type: Successful
Phase: Successful
The operator evicts the secondary (hanadb-cluster-1) first and the primary (hanadb-cluster-0) last.
The pods now show a fresh age and the database is back to Ready. Note that restarting the old primary
triggers a normal HANA SystemReplication takeover, so the primary/secondary roles may swap:
$ kubectl get pods -n demo -l app.kubernetes.io/instance=hanadb-cluster -L kubedb.com/role
NAME READY STATUS RESTARTS AGE ROLE
hanadb-cluster-0 2/2 Running 0 4m30s secondary
hanadb-cluster-1 2/2 Running 0 7m16s primary
hanadb-cluster-arbiter-0 1/1 Running 0 15m arbiter
$ kubectl get hanadb.kubedb.com -n demo hanadb-cluster
NAME VERSION STATUS AGE
hanadb-cluster 2.0.82 Ready 22m
Cleaning Up
$ kubectl delete hdbops -n demo hdbops-restart
$ kubectl delete hanadb.kubedb.com -n demo hanadb-cluster
$ kubectl delete ns demo
Next Steps
- Vertically scale a HanaDB.
- Review the HanaDBOpsRequest CRD.