KubeDB

By AppsCode

CONTACT SALES

New to KubeDB? Please start here.

Horizontal Scale Distributed MariaDB

This guide will show you how to use KubeDB Opsrequest operator to horizontally scale a distributed MariaDB Galera cluster across multiple Kubernetes clusters.

Note: All other OpsRequest operations behave consistently in a distributed environment. Only HorizontalScaling has cluster-specific considerations covered in this guide.

Before You Begin

To keep everything isolated, we are going to use a separate namespace called demo throughout this tutorial.

$ kubectl create ns demo
namespace/demo created

Apply Horizontal Scaling on Distributed Cluster

Here, we are going to deploy a distributed MariaDB Galera cluster and then apply horizontal scaling using MariaDBOpsRequest.

Deploy PlacementPolicy

For a distributed MariaDB cluster, the PlacementPolicy must be created before deploying the database. It defines which replica index is scheduled on which cluster and acts as the upper bound for scaling, the actual running replicas can be any number less than or equal to the total indices defined in the policy.

In this example, the PlacementPolicy is configured with five replica indices distributed across two clusters. The MariaDB cluster will start with three replicas (indices 0, 1, 2), and can be scaled up to five (indices 0–4) without modifying the policy.

apiVersion: apps.k8s.appscode.com/v1
kind: PlacementPolicy
metadata:
  labels:
    app.kubernetes.io/managed-by: Helm
  name: distributed-mariadb
spec:
  clusterSpreadConstraint:
    distributionRules:
      - clusterName: demo-controller
        replicaIndices:
          - 0
          - 2
          - 4
      - clusterName: demo-worker
        replicaIndices:
          - 1
          - 3
    slice:
      projectNamespace: kubeslice-demo-distributed-mariadb
      sliceName: demo-slice
  nodeSpreadConstraint:
    maxSkew: 1
    whenUnsatisfiable: ScheduleAnyway
  zoneSpreadConstraint:
    maxSkew: 1
    whenUnsatisfiable: ScheduleAnyway

Here,

  • spec.clusterSpreadConstraint.distributionRules[].replicaIndices specifies which MariaDB replica indices are scheduled on that cluster. demo-controller will host replicas 0, 2, and 4; demo-worker will host replicas 1 and 3.

Apply the PlacementPolicy on the hub (demo-controller) cluster:

$ kubectl apply -f https://github.com/kubedb/docs/raw/v2026.4.27/docs/guides/mariadb/distributed/opsrequest/examples/placement-policy.yaml --context demo-controller
placementpolicy.apps.k8s.appscode.com/distributed-mariadb created

Deploy Distributed MariaDB Cluster

In this section, we are going to deploy a distributed MariaDB Galera cluster with version 11.5.2. Note that spec.distributed is set to true and spec.replicas is 3 which is less than the five indices defined in the PlacementPolicy. The operator will only create pods for indices 0, 1, and 2.

apiVersion: kubedb.com/v1
kind: MariaDB
metadata:
  name: mariadb
  namespace: demo
spec:
  version: "11.5.2"
  distributed: true
  replicas: 3
  storageType: Durable
  storage:
    storageClassName: "longhorn"
    accessModes:
      - ReadWriteOnce
    resources:
      requests:
        storage: 1Gi
  podTemplate:
    spec:
      podPlacementPolicy:
        name: distributed-mariadb
      containers:
        - name: mariadb
          resources:
            requests:
              cpu: "200m"
              memory: "300Mi"
            limits:
              cpu: "200m"
              memory: "300Mi"
  deletionPolicy: WipeOut

Let’s create the MariaDB CR we have shown above,

$ kubectl create -f https://github.com/kubedb/docs/raw/v2026.4.27/docs/guides/mariadb/distributed/opsrequest/examples/mariadb.yaml --context demo-controller
mariadb.kubedb.com/mariadb created

Now, wait until mariadb has status Ready:

$ kubectl get mariadb -n demo --context demo-controller
NAME             VERSION   STATUS   AGE
mariadb          11.5.2    Ready    2m36s

Let’s check the number of replicas this database has from the MariaDB object:

$ kubectl get mariadb -n demo mariadb -o json | jq '.spec.replicas'
3

The pods are distributed across clusters as defined by the PlacementPolicy. Indices 0 and 2 land on demo-controller; index 1 lands on demo-worker:

$ kubectl get pods -n demo --context demo-controller
NAME          READY   STATUS    RESTARTS   AGE
mariadb-0     3/3     Running   0          2m30s
mariadb-2     3/3     Running   0          2m30s

$ kubectl get pods -n demo --context demo-worker
NAME          READY   STATUS    RESTARTS   AGE
mariadb-1     3/3     Running   0          2m30s

We can see the cluster has 3 pods. We are now ready to apply MariaDBOpsRequest to scale this database.

Scale Up Replicas

Here, we are going to scale up the replicas from 3 to 5. Because the PlacementPolicy already defines indices 0–4, no changes to the policy are required — the operator knows where to place the new pods.

Create MariaDBOpsRequest

In order to scale up the replicas, we have to create a MariaDBOpsRequest CR with our desired replicas. Below is the YAML of the MariaDBOpsRequest CR that we are going to create:

apiVersion: ops.kubedb.com/v1alpha1
kind: MariaDBOpsRequest
metadata:
  name: mdops-scale-horizontal-up
  namespace: demo
spec:
  type: HorizontalScaling
  databaseRef:
    name: mariadb
  horizontalScaling:
    member : 5

Here,

  • spec.databaseRef.name specifies that we are performing horizontal scaling on mariadb.
  • spec.type specifies that we are performing HorizontalScaling.
  • spec.horizontalScaling.member specifies the desired replica count after scaling. This value must not exceed the total number of indices defined in the PlacementPolicy (5 in this example).

Let’s create the MariaDBOpsRequest CR we have shown above,

$ kubectl apply -f https://github.com/kubedb/docs/raw/v2026.4.27/docs/guides/mariadb/distributed/opsrequest/examples/mdops-upscale.yaml --context demo-controller
mariadbopsrequest.ops.kubedb.com/mdops-scale-horizontal-up created

Verify Cluster Replicas Scaled Up Successfully

If everything goes well, KubeDB Enterprise operator will update the replicas of the MariaDB object and create the new pods on the clusters specified by the PlacementPolicy.

Let’s wait for MariaDBOpsRequest to be Successful:

$ watch kubectl get mariadbopsrequest -n demo --context demo-controller
Every 2.0s: kubectl get mariadbopsrequest -n demo
NAME                        TYPE                STATUS       AGE
mdops-scale-horizontal-up   HorizontalScaling   Successful   18m

We can see from the above output that the MariaDBOpsRequest has succeeded. Now, let’s verify the number of replicas:

$ kubectl get mariadb -n demo mariadb -o json | jq '.spec.replicas'
5

The two new pods are placed on the clusters according to the PlacementPolicy — index 4 on demo-controller and index 3 on demo-worker:

$ kubectl get pods -n demo --context demo-controller
NAME        READY   STATUS    RESTARTS   AGE
mariadb-0   3/3     Running   0          58m
mariadb-2   3/3     Running   0          55m
mariadb-4   3/3     Running   0          17m


$ kubectl get pods -n demo --context demo-worker
NAME        READY   STATUS    RESTARTS   AGE
mariadb-1   3/3     Running   0          57m
mariadb-3   3/3     Running   0          19m

From all the above outputs we can see that the cluster now has 5 replicas. We have successfully scaled up the distributed MariaDB cluster.

Scale Down Replicas

Here, we are going to scale down the replicas from 5 to 3. The PlacementPolicy does not need to be modified — the operator will remove the highest-indexed pods (mariadb-3 and mariadb-4) and the remaining pods will continue running as defined by the policy.

Create MariaDBOpsRequest

In order to scale down the cluster, we have to create a MariaDBOpsRequest CR with our desired replicas. Below is the YAML of the MariaDBOpsRequest CR that we are going to create:

apiVersion: ops.kubedb.com/v1alpha1
kind: MariaDBOpsRequest
metadata:
  name: mdops-scale-horizontal-down
  namespace: demo
spec:
  type: HorizontalScaling
  databaseRef:
    name: mariadb
  horizontalScaling:
    member: 3

Here,

  • spec.databaseRef.name specifies that we are performing horizontal scaling down on mariadb.
  • spec.type specifies that we are performing HorizontalScaling.
  • spec.horizontalScaling.member specifies the desired replica count after scaling.

Let’s create the MariaDBOpsRequest CR we have shown above,

$ kubectl apply -f https://github.com/kubedb/docs/raw/v2026.4.27/docs/guides/mariadb/distributed/opsrequest/examples/mdops-downscale.yaml --context demo-controller
mariadbopsrequest.ops.kubedb.com/mdops-scale-horizontal-down created

Verify Cluster Replicas Scaled Down Successfully

If everything goes well, KubeDB Enterprise operator will update the replicas of the MariaDB object and remove the highest-indexed pods from their respective clusters.

Let’s wait for MariaDBOpsRequest to be Successful:

$ watch kubectl get mariadbopsrequest -n demo --context demo-controller
Every 2.0s: kubectl get mariadbopsrequest -n demo
NAME                          TYPE               STATUS       AGE
mdops-scale-horizontal-down   HorizontalScaling  Successful   2m32s

We can see from the above output that the MariaDBOpsRequest has succeeded. Now, let’s verify the number of replicas:

$ kubectl get mariadb -n demo mariadb -o json | jq '.spec.replicas'
3

Pods mariadb-3 and mariadb-4 have been removed from their respective clusters:

$ kubectl get pods -n demo --context demo-controller
NAME          READY   STATUS    RESTARTS   AGE
mariadb-0     3/3     Running   0          20m
mariadb-2     3/3     Running   0          20m

$ kubectl get pods -n demo --context demo-worker
NAME          READY   STATUS    RESTARTS   AGE
mariadb-1     3/3     Running   0          20m

From all the above outputs we can see that the cluster now has 3 replicas. We have successfully scaled down the distributed MariaDB cluster.

Cleaning Up

To clean up the Kubernetes resources created by this tutorial, run:

$ kubectl delete mariadb -n demo mariadb --context demo-controller
$ kubectl delete mariadbopsrequest -n demo mdops-scale-horizontal-up mdops-scale-horizontal-down --context demo-controller
$ kubectl delete placementpolicy distributed-mariadb --context demo-controller

What's on this Page

Search
Improve This Page