redis-cluster-operator

README

redis-cluster-operator

Overview

Redis Cluster Operator manages Redis Cluster atop Kubernetes.

The operator itself is built with the Operator framework.

Each master node and its slave nodes is managed by a statefulSet, create a headless svc for each statefulSet, and create a clusterIP service for all nodes.

Each statefulset uses PodAntiAffinity to ensure that the master and slaves are dispersed on different nodes. At the same time, when the operator selects the master in each statefulset, it preferentially select the pod with different k8s nodes as master.

Table of Contents

• redis-cluster-operator
  • Overview
  • Prerequisites
  • Features
  • Quick Start
    • Deploy redis cluster operator
      • Deploy a sample Redis Cluster
      • Scaling Up the Redis Cluster
      • Scaling Down the Redis Cluster
      • Backup and Restore
      • Prometheus Discovery
      • Create Redis Cluster with password
      • Persistent Volume
      • Custom Configuration
      • Custom Service
      • Custom Resource
  • ValidatingWebhook
  • End to end tests

Prerequisites

• go version v1.13+.
• Access to a Kubernetes v1.13.10 cluster.

Features

• Customize the number of master nodes and the number of replica nodes per master

• Password

• Safely Scaling the Redis Cluster

• Backup and Restore

• Persistent Volume

• Custom Configuration

• Prometheus Discovery

Quick Start

Deploy redis cluster operator

Register the DistributedRedisCluster and RedisClusterBackup custom resource definition (CRD).

$ kubectl create -f deploy/crds/redis.kun_distributedredisclusters_crd.yaml
$ kubectl create -f deploy/crds/redis.kun_redisclusterbackups_crd.yaml

A namespace-scoped operator watches and manages resources in a single namespace, whereas a cluster-scoped operator watches and manages resources cluster-wide. You can chose run your operator as namespace-scoped or cluster-scoped.

// cluster-scoped
$ kubectl create -f deploy/service_account.yaml
$ kubectl create -f deploy/cluster/cluster_role.yaml
$ kubectl create -f deploy/cluster/cluster_role_binding.yaml
$ kubectl create -f deploy/cluster/operator.yaml

// namespace-scoped
$ kubectl create -f deploy/service_account.yaml
$ kubectl create -f deploy/namespace/role.yaml
$ kubectl create -f deploy/namespace/role_binding.yaml
$ kubectl create -f deploy/namespace/operator.yaml

Verify that the redis-cluster-operator is up and running:

$ kubectl get deployment
NAME                     READY   UP-TO-DATE   AVAILABLE   AGE
redis-cluster-operator   1/1     1            1           1d

Deploy a sample Redis Cluster

$ kubectl apply -f deploy/example/redis.kun_v1alpha1_distributedrediscluster_cr.yaml

Verify that the cluster instances and its components are running.

$ kubectl get distributedrediscluster
NAME                              MASTERSIZE   STATUS    AGE
example-distributedrediscluster   3            Scaling   11s

$ kubectl get all -l redis.kun/name=example-distributedrediscluster
NAME                                          READY   STATUS    RESTARTS   AGE
pod/drc-example-distributedrediscluster-0-0   1/1     Running   0          2m48s
pod/drc-example-distributedrediscluster-0-1   1/1     Running   0          2m8s
pod/drc-example-distributedrediscluster-1-0   1/1     Running   0          2m48s
pod/drc-example-distributedrediscluster-1-1   1/1     Running   0          2m13s
pod/drc-example-distributedrediscluster-2-0   1/1     Running   0          2m48s
pod/drc-example-distributedrediscluster-2-1   1/1     Running   0          2m15s

NAME                                        TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)              AGE
service/example-distributedrediscluster     ClusterIP   172.17.132.71   <none>        6379/TCP,16379/TCP   2m48s
service/example-distributedrediscluster-0   ClusterIP   None            <none>        6379/TCP,16379/TCP   2m48s
service/example-distributedrediscluster-1   ClusterIP   None            <none>        6379/TCP,16379/TCP   2m48s
service/example-distributedrediscluster-2   ClusterIP   None            <none>        6379/TCP,16379/TCP   2m48s

NAME                                                     READY   AGE
statefulset.apps/drc-example-distributedrediscluster-0   2/2     2m48s
statefulset.apps/drc-example-distributedrediscluster-1   2/2     2m48s
statefulset.apps/drc-example-distributedrediscluster-2   2/2     2m48s

$ kubectl get distributedrediscluster
NAME                              MASTERSIZE   STATUS    AGE
example-distributedrediscluster   3            Healthy   4m

Scaling Up the Redis Cluster

Increase the masterSize to trigger the scaling up.

apiVersion: redis.kun/v1alpha1
kind: DistributedRedisCluster
metadata:
  annotations:
    # if your operator run as cluster-scoped, add this annotations
    redis.kun/scope: cluster-scoped
  name: example-distributedrediscluster
spec:
  # Increase the masterSize to trigger the scaling.
  masterSize: 4
  ClusterReplicas: 1
  image: redis:5.0.4-alpine

Scaling Down the Redis Cluster

Decrease the masterSize to trigger the scaling down.

apiVersion: redis.kun/v1alpha1
kind: DistributedRedisCluster
metadata:
  annotations:
    # if your operator run as cluster-scoped, add this annotations
    redis.kun/scope: cluster-scoped
  name: example-distributedrediscluster
spec:
  # Increase the masterSize to trigger the scaling.
  masterSize: 3
  ClusterReplicas: 1
  image: redis:5.0.4-alpine

Backup and Restore

Only Ceph object storage is supported now

Backup

$ kubectl create -f deploy/example/backup-restore/redisclusterbackup_cr.yaml

Restore from backup

$ kubectl create -f deploy/example/backup-restore/restore.yaml

Prometheus Discovery

$ kubectl create -f deploy/example/prometheus-exporter.yaml

Create Redis Cluster with password

$ kubectl create -f deploy/example/custom-password.yaml

Persistent Volume

$ kubectl create -f deploy/example/persistent.yaml

Custom Configuration

$ kubectl create -f deploy/example/custom-config.yaml

Custom Service

$ kubectl create -f deploy/example/custom-service.yaml

Custom Resource

$ kubectl create -f deploy/example/custom-resources.yaml

ValidatingWebhook

see ValidatingWebhook

End to end tests

see e2e