Skip to content

clusters.md

Latest commit

 

History

History
203 lines (173 loc) · 4.48 KB

clusters.md

File metadata and controls

203 lines (173 loc) · 4.48 KB

Clusters

MySQL cluster examples.

Create a cluster with 3 members

The following example will create a MySQL Cluster with 3 members, one primary and 2 secondaries:

apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-test-cluster
spec:
  members: 3

Create a cluster with 3 members in multi-master mode

The following example will create a MySQL Cluster with 3 primary (read/write) members:

apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-multimaster-cluster
spec:
  multiMaster: true
  members: 3

Create a cluster with a custom "MYSQL_ROOT_PASSWORD"

Create your own secret with a password field

$ kubectl create secret generic mysql-root-user-secret --from-literal=password=foobar

Create your cluster and reference it

apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-cluster-custom-secret
spec:
  members: 1
  rootPasswordSecret:
    name: mysql-root-user-secret

Create a cluster with a persistent volume

The following example will create a MySQL Cluster with a persistent local volume.

---
apiVersion: v1
kind: PersistentVolume
metadata:
  labels:
    type: local
  name: mysql-local-volume
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  hostPath:
    path: /tmp/data
  persistentVolumeReclaimPolicy: Recycle
  storageClassName: manual
---
apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-cluster-with-volume
spec:
  members: 1
  volumeClaimTemplate:
    metadata:
      name: data
    spec:
      storageClassName: manual
      accessModes:
        - ReadWriteMany
      resources:
        requests:
          storage: 1Gi

Create a cluster with a persistent data volume and a persistent volume to use for backups/restore

The following example will create a MySQL Cluster with a persistent local data volume and a persistent local backup/restore volume.

---
apiVersion: v1
kind: PersistentVolume
metadata:
  labels:
    type: local
  name: mysql-local-volume
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  hostPath:
    path: /tmp/data1
  persistentVolumeReclaimPolicy: Recycle
  storageClassName: manual
---
apiVersion: v1
kind: PersistentVolume
metadata:
  labels:
    type: local
  name: mysql-local-backup-volume
spec:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  hostPath:
    path: /tmp/data2
  persistentVolumeReclaimPolicy: Recycle
  storageClassName: manual
---
apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-cluster-with-volume
spec:
  members: 1
  rootPasswordSecret:
    name: mysql-root-user-secret
  volumeClaimTemplate:
    metadata:
      name: data
    spec:
      storageClassName: manual
      accessModes:
        - ReadWriteMany
      resources:
        requests:
          storage: 1Gi
  backupVolumeClaimTemplate:
    metadata:
      name: backup-data
    spec:
      storageClassName: manual
      accessModes:
        - ReadWriteMany
      resources:
        requests:
          storage: 1Gi

Create a cluster with custom my.cnf configuration file

By default, the MySQL Operator starts a cluster with an opinionated set of defaults.

However, you may wish to configure some aspects of your cluster through a my.cnf config file. This can be achieved by creating a config map and referencing it as part of your cluster spec.

Create the config map

First we create a config map containing the configuration file we want to apply to our cluster.

kubectl create configmap mycnf --from-file=examples/my.cnf

Reference it in the cluster spec

Now we can reference our config map in our cluster spec definition. For example:

apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-cluster-with-config
  members: 3
  config:
    name: mycnf

Create a cluster with custom server_id values

By default, the MySQL Operator starts a cluster with --server_id set to 1000 and increments it by one for each new cluster member. You can change this behavior by setting the baseServerId field on your Cluster. baseServerId value can be set to anything in the range from 1 to 4294967286. 0 is also accepted, but then the default value of 1000 will be used.

The following example will create a MySQL Cluster with following server_id's: 42,43,44

apiVersion: mysql.oracle.com/v1alpha1
kind: Cluster
metadata:
  name: mysql-cluster-with-custom-serverid
  members: 3
  baseServerId: 42