kubernetes redis-cluster生产实践笔记

Z时代
2024-01-10
分类：技术分享

方案一自定义yaml文件安装redis cluster

背景

在Kubernetes中部署Redis集群面临挑战，因为每个Redis实例都依赖于一个配置文件，该文件可以跟踪其他集群实例及其角色。为此，需要结合使用Kubernetes StatefulSets和PersistentVolumes实现。

redis cluster架构示意图：

kubernetes生产实践之redis-cluster

创建StatefulSet yaml文件

--- apiVersion: v1 kind: ConfigMap metadata: name: redis-cluster data: update.sh: | #!/bin/sh REDIS_NODES="/data/nodes.conf" sed -i -e "/myself/ s/[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}/${POD_IP}/" ${REDIS_NODES} exec "[email protected]" redis.conf: |+ bind 0.0.0.0 cluster-enabled yes cluster-require-full-coverage no cluster-node-timeout 30000 cluster-config-file /data/nodes.conf cluster-migration-barrier 1 appendonly yes protected-mode no --- apiVersion: apps.kruise.io/v1beta1 # apiVersion: apps/v1 kind: StatefulSet metadata: name: redis-cluster spec: serviceName: redis-cluster replicas: 6 selector: matchLabels: app: redis-cluster template: metadata: labels: app: redis-cluster spec: containers: - name: redis image: redis:6.2.1-alpine ports: - containerPort: 6379 name: client - containerPort: 16379 name: gossip command: ["/conf/update.sh", "redis-server", "/conf/redis.conf"] env: - name: POD_IP valueFrom: fieldRef: fieldPath: status.podIP volumeMounts: - name: conf mountPath: /conf readOnly: false - name: data mountPath: /data readOnly: false volumes: - name: conf configMap: name: redis-cluster defaultMode: 0755 volumeClaimTemplates: - metadata: name: data spec: accessModes: [ "ReadWriteOnce" ] resources: requests: storage: 100Gi storageClassName: rbd

注释：

cluster-migration-barrier 那些分配后仍然剩余migration barrier个从节点的主节点才会触发节点分配，而不是分配前有migration barrier个从节点的主节点就会触发节点分配，默认是1，生产环境建议维持默认值 protected-mode no 参数是为了禁止外网访问redis，如果启用了，则只能够通过lookback ip（127.0.0.1）访问Redis，如果外网访问redis，会报出异常。 apiVersion: apps.kruise.io/v1beta1 控制器这里使用kruise提供的Advanced StatefulSet，如果集群没有安装kruise，可以使用 apps/v1

安装redis-cluster

[[email protected] redis]# kubectl apply -f install-redis.yaml configmap/redis-cluster created statefulset.apps.kruise.io/redis-cluster created [[email protected] redis]# kubectl get po -n op NAME READY STATUS RESTARTS AGE redis-cluster-0 1/1 Running 0 3m26s redis-cluster-1 1/1 Running 0 3m14s redis-cluster-2 1/1 Running 0 2m54s redis-cluster-3 1/1 Running 0 2m23s redis-cluster-4 1/1 Running 0 2m14s redis-cluster-5 1/1 Running 0 114s

创建redis-cluster service

--- apiVersion: v1 kind: Service metadata: name: redis-cluster namespace: op spec: type: ClusterIP ports: - port: 6379 targetPort: 6379 name: client - port: 16379 targetPort: 16379 name: gossip selector: app: redis-cluster

[[email protected] redis]# kubectl apply -f redis-svc.yml
service/redis-cluster created
[[email protected] redis]# kubectl get svc -n op
NAME            TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)              AGE
redis-cluster   ClusterIP   10.97.197.224   <none>        6379/TCP,16379/TCP   9s
测试能后连通
[[email protected] redis]# telnet  10.97.197.224 6379
Trying 10.97.197.224...
Connected to 10.97.197.224.
Escape character is '^]'.

初始化redis-cluster

执行如下命令，获取到pod IP，然后使用redis-cli --cluster创建集群

[[email protected] redis]# kubectl -n op exec -it redis-cluster-0 -- redis-cli --cluster create --cluster-replicas 1 $(kubectl -n op get pods -l app=redis-cluster -o jsonpath='{range.items[*]}{.status.podIP}:6379 {end}')
>>> Performing hash slots allocation on 6 nodes...
Master[0] -> Slots 0 - 5460
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
Adding replica 100.88.43.67:6379 to 100.64.147.152:6379
Adding replica 100.113.170.5:6379 to 100.98.174.217:6379
Adding replica 100.64.147.153:6379 to 100.80.158.227:6379
M: b47b27a3dbddf3fc1370cbe14ae753f4fce20b04 100.64.147.152:6379
slots:[0-5460] (5461 slots) master
M: 09543217c903350e963fc4fdf4acb73f8a1b7f8b 100.98.174.217:6379
slots:[5461-10922] (5462 slots) master
M: 5389ace495b68eeac85370d6783648dff68f2fb6 100.80.158.227:6379
slots:[10923-16383] (5461 slots) master
S: b1f39714c006ae55b12b18e6537303d7a00e1704 100.64.147.153:6379
replicates 5389ace495b68eeac85370d6783648dff68f2fb6
S: 0113f4668ec2f3ca2e9470c44bd5faab532b0936 100.88.43.67:6379
replicates b47b27a3dbddf3fc1370cbe14ae753f4fce20b04
S: e1e2f18ae66c79f1943390beabb59613abbad38a 100.113.170.5:6379
replicates 09543217c903350e963fc4fdf4acb73f8a1b7f8b
Can I set the above configuration? (type 'yes' to accept): yes
>>> Nodes configuration updated
>>> Assign a different config epoch to each node
>>> Sending CLUSTER MEET messages to join the cluster
Waiting for the cluster to join
..
>>> Performing Cluster Check (using node 100.64.147.152:6379)
M: b47b27a3dbddf3fc1370cbe14ae753f4fce20b04 100.64.147.152:6379
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: 0113f4668ec2f3ca2e9470c44bd5faab532b0936 100.88.43.67:6379
slots: (0 slots) slave
replicates b47b27a3dbddf3fc1370cbe14ae753f4fce20b04
M: 09543217c903350e963fc4fdf4acb73f8a1b7f8b 100.98.174.217:6379
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
M: 5389ace495b68eeac85370d6783648dff68f2fb6 100.80.158.227:6379
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: e1e2f18ae66c79f1943390beabb59613abbad38a 100.113.170.5:6379
slots: (0 slots) slave
replicates 09543217c903350e963fc4fdf4acb73f8a1b7f8b
S: b1f39714c006ae55b12b18e6537303d7a00e1704 100.64.147.153:6379
slots: (0 slots) slave
replicates 5389ace495b68eeac85370d6783648dff68f2fb6
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

验证集群信息

[[email protected] redis]# kubectl -n op exec -it redis-cluster-0 -- redis-cli cluster info cluster_state:ok cluster_slots_assigned:16384 cluster_slots_ok:16384 cluster_slots_pfail:0 cluster_slots_fail:0 cluster_known_nodes:6 cluster_size:3 cluster_current_epoch:6 cluster_my_epoch:1 cluster_stats_messages_ping_sent:178 cluster_stats_messages_pong_sent:181 cluster_stats_messages_sent:359 cluster_stats_messages_ping_received:176 cluster_stats_messages_pong_received:178 cluster_stats_messages_meet_received:5 cluster_stats_messages_received:359 [[email protected] redis]# kubectl -n op exec -it redis-cluster-0 -- redis-cli cluster nodes 0113f4668ec2f3ca2e9470c44bd5faab532b0936 100.88.43.67:[email protected] slave b47b27a3dbddf3fc1370cbe14ae753f4fce20b04 0 1615348311156 1 connected 09543217c903350e963fc4fdf4acb73f8a1b7f8b 100.98.174.217:[email protected] master - 0 1615348314162 2 connected 5461-10922 b47b27a3dbddf3fc1370cbe14ae753f4fce20b04 100.64.147.152:[email protected] myself,master - 0 1615348312000 1 connected 0-5460 5389ace495b68eeac85370d6783648dff68f2fb6 100.80.158.227:[email protected] master - 0 1615348312000 3 connected 10923-16383 e1e2f18ae66c79f1943390beabb59613abbad38a 100.113.170.5:[email protected] slave 09543217c903350e963fc4fdf4acb73f8a1b7f8b 0 1615348313160 2 connected b1f39714c006ae55b12b18e6537303d7a00e1704 100.64.147.153:[email protected] slave 5389ace495b68eeac85370d6783648dff68f2fb6 0 1615348312158 3 connected

从输出可以看到，集群总共6个节点，三主三从

方案二使用kubeDB安装redis

安装kubeDB

1、安装KubeDB

获取AppsCode License https://license-issuer.appscode.com/

下载KubeDB charts https://github.com/appscode/charts/tree/master/stable/kubedb-community

[[email protected] kubedb-community]# unzip kubedb-community-v0.16.2.tgz [[email protected] kubedb-community]# cd kubedb-community [[email protected] kubedb-community]# ls -al total 96 drwxr-xr-x 4 root root 158 Mar 10 15:26 . drwxr-xr-x 3 root root 66 Mar 10 15:24 .. -rw-r--r-- 1 root root 351 Feb 16 09:55 Chart.yaml drwxr-xr-x 2 root root 28 Mar 10 15:24 ci -rw-r--r-- 1 root root 493 Feb 16 09:55 doc.yaml -rw-r--r-- 1 root root 353 Feb 16 09:55 .helmignore -rw-r--r-- 1 root root 24422 Feb 16 09:55 README.md drwxr-xr-x 2 root root 4096 Mar 10 15:24 templates -rw-r--r-- 1 root root 47437 Feb 16 09:55 values.openapiv3_schema.yaml -rw-r--r-- 1 root root 5230 Feb 16 09:55 values.yaml

修改values.yaml，把License文件放到kubedb-community目录下

2、使用helm安装

[[email protected] kubedb-community]# helm install kubedb-community --namespace kube-system --set-file license=./kubedb-community-license.txt -f values.yaml  .
NAME: kubedb-community
LAST DEPLOYED: Wed Mar 10 15:38:59 2021
NAMESPACE: kube-system
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
To verify that KubeDB has started, run:
kubectl get deployment --namespace kube-system -l "app.kubernetes.io/name=kubedb-community,app.kubernetes.io/instance=kubedb-community"
Now install/upgrade appscode/kubedb-catalog chart.
To install, run:
helm install kubedb-catalog appscode/kubedb-catalog --version v0.16.2 --namespace kube-system
To upgrade, run:
helm upgrade kubedb-catalog appscode/kubedb-catalog --version v0.16.2 --namespace kube-system
执行如下命令查看是否安装完成
[[email protected] kubedb-community]# kubectl get deployment --namespace kube-system -l "app.kubernetes.io/name=kubedb-community,app.kubernetes.io/instance=kubedb-community"
NAME               READY   UP-TO-DATE   AVAILABLE   AGE
kubedb-community   1/1     1            1           38s
等待crds注册成功
[[email protected] kubedb-community]# kubectl get crds -l app.kubernetes.io/name=kubedb -w
NAME                                       CREATED AT
elasticsearches.kubedb.com                 2021-03-10T07:39:42Z
elasticsearchversions.catalog.kubedb.com   2021-03-10T07:39:45Z
etcds.kubedb.com                           2021-03-10T07:39:42Z
etcdversions.catalog.kubedb.com            2021-03-10T07:39:45Z
memcacheds.kubedb.com                      2021-03-10T07:39:43Z
memcachedversions.catalog.kubedb.com       2021-03-10T07:39:45Z
mongodbs.kubedb.com                        2021-03-10T07:39:43Z
mongodbversions.catalog.kubedb.com         2021-03-10T07:39:45Z
mysqls.kubedb.com                          2021-03-10T07:39:43Z
mysqlversions.catalog.kubedb.com           2021-03-10T07:39:46Z
perconaxtradbs.kubedb.com                  2021-03-10T07:39:43Z
perconaxtradbversions.catalog.kubedb.com   2021-03-10T07:39:46Z
pgbouncers.kubedb.com                      2021-03-10T07:39:44Z
pgbouncerversions.catalog.kubedb.com       2021-03-10T07:39:46Z
postgreses.kubedb.com                      2021-03-10T07:39:44Z
postgresversions.catalog.kubedb.com        2021-03-10T07:39:46Z
proxysqls.kubedb.com                       2021-03-10T07:39:44Z
proxysqlversions.catalog.kubedb.com        2021-03-10T07:39:46Z
redises.kubedb.com                         2021-03-10T07:39:45Z
redisversions.catalog.kubedb.com           2021-03-10T07:39:46Z

3、安装KubeDB Catalog

同样，先下载https://github.com/appscode/charts/tree/master/stable/kubedb-catalog

[[email protected] kubedb-catalog]# tar -zxf kubedb-catalog-v0.16.2.tgz
[[email protected] kubedb-catalog]# cd kubedb-catalog
[[email protected] kubedb-catalog]# ls -al
total 24
drwxr-xr-x  3 root root  148 Mar 10 15:48 .
drwxr-xr-x  3 root root   28 Mar 10 15:48 ..
-rw-r--r--  1 root root  321 Jan 26 20:08 Chart.yaml
-rw-r--r--  1 root root  467 Jan 26 20:08 doc.yaml
-rw-r--r--  1 root root  353 Jan 26 20:08 .helmignore
-rw-r--r--  1 root root 3195 Jan 26 20:08 README.md
drwxr-xr-x 12 root root  188 Mar 10 15:48 templates
-rw-r--r--  1 root root  744 Jan 26 20:08 values.openapiv3_schema.yaml
-rw-r--r--  1 root root 1070 Jan 26 20:08 values.yaml
[[email protected] kubedb-catalog]# helm install kubedb-catalog --namespace kube-system -f values.yaml  .
NAME: kubedb-catalog
LAST DEPLOYED: Wed Mar 10 15:50:50 2021
NAMESPACE: kube-system
STATUS: deployed
REVISION: 1
TEST SUITE: None

使用kubedb安装redis

1、先看下官方给的redis生命周期示意图

kubernetes生产实践之redis-cluster

kubedb安装redis支持如下特性

Features Availability Clustering ✓ Instant Backup ✗ Scheduled Backup ✗ Persistent Volume ✓ Initialize using Snapshot ✗ Initialize using Script ✗ Custom Configuration ✓ Using Custom docker image ✓ Builtin Prometheus Discovery ✓ Using Prometheus operator ✓

2、查看支持的版本

[[email protected] kubedb-catalog]# kubectl get redisversions NAME VERSION DB_IMAGE DEPRECATED AGE 4.0.11 4.0.11 kubedb/redis:4.0.11 15m 4.0.6-v2 4.0.6 kubedb/redis:4.0.6-v2 15m 5.0.3-v1 5.0.3 kubedb/redis:5.0.3-v1 15m 6.0.6 6.0.6 kubedb/redis:6.0.6 15m

3、编辑yaml安装文件

可以参照https://github.com/kubedb/docs/blob/v2021.01.26/docs/examples/redis/clustering/demo-1.yaml

这里选择安装6.0.6这个版本，我的集群storageClassName: "rbd"，请根据实际修改

如果想自定义redis.conf,请参考https://github.com/kubedb/docs/blob/v2021.01.26/docs/examples/redis/custom-config/redis-custom.yaml

apiVersion: kubedb.com/v1alpha2 kind: Redis metadata: name: redis-cluster namespace: op spec: version: 6.0.6 mode: Cluster cluster: master: 3 replicas: 1 storageType: Durable storage: resources: requests: storage: 1Gi storageClassName: "rbd" accessModes: - ReadWriteOnce

执行安装

[[email protected] kubedb-community]# kubectl apply -f redis-cluster.yaml redis.kubedb.com/redis-cluster created 安装完成，可以如下查看 [[email protected] kubedb-community]# kubectl get rd,po -n op NAME VERSION STATUS AGE redis.kubedb.com/redis-cluster 6.0.6 Provisioning 6m55s NAME READY STATUS RESTARTS AGE pod/redis-cluster-shard0-0 1/1 Running 0 6m54s pod/redis-cluster-shard0-1 1/1 Running 0 6m18s pod/redis-cluster-shard1-0 1/1 Running 0 5m38s pod/redis-cluster-shard1-1 1/1 Running 0 5m1s pod/redis-cluster-shard2-0 1/1 Running 0 4m30s pod/redis-cluster-shard2-1 1/1 Running 0 4m8s [[email protected] redis]# kubectl get svc -n op NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE redis-cluster ClusterIP 10.97.197.224 <none> 6379/TCP 5h16m redis-cluster-pods ClusterIP None <none> 6379/TCP 17m

4、验证集群

[[email protected] kubedb-community]# kubectl get pods -n op -o jsonpath='{range.items[*]}{.metadata.name} ---------- {.status.podIP}:6379{"\t\n"}{end}' | grep redis redis-cluster-shard0-0 ---------- 100.64.147.156:6379 redis-cluster-shard0-1 ---------- 100.98.174.218:6379 redis-cluster-shard1-0 ---------- 100.126.252.204:6379 redis-cluster-shard1-1 ---------- 100.113.170.6:6379 redis-cluster-shard2-0 ---------- 100.107.55.69:6379 redis-cluster-shard2-1 ---------- 100.78.230.4:6379 [[email protected] redis]# kubectl -n op exec -it redis-cluster-shard0-0 -- redis-cli cluster info cluster_state:ok cluster_slots_assigned:16384 cluster_slots_ok:16384 cluster_slots_pfail:0 cluster_slots_fail:0 cluster_known_nodes:6 cluster_size:3 cluster_current_epoch:3 cluster_my_epoch:1 cluster_stats_messages_ping_sent:864 cluster_stats_messages_pong_sent:882 cluster_stats_messages_sent:1746 cluster_stats_messages_ping_received:879 cluster_stats_messages_pong_received:864 cluster_stats_messages_meet_received:3 cluster_stats_messages_received:1746 [[email protected] redis]# kubectl -n op exec -it redis-cluster-shard0-0 -- redis-cli cluster nodes 1895cb4b9c31b848666c61000e502f55a29a8255 100.64.147.155:[email protected] master - 0 1615365162008 2 connected 5461-10922 30bdbf2ca37001774498a9b935afbc1cd2ce389c 100.126.252.203:[email protected] slave 2c06092fafa99e0158e39e6237a04fed25be3550 0 1615365163000 1 connected 9b2cfbd5c1b417121d410141b6da9512ad29ce3c 100.78.230.3:[email protected] slave e83446c368839c5fdccf5f70e3b1004eb67cb651 0 1615365163512 3 connected 2c06092fafa99e0158e39e6237a04fed25be3550 100.82.197.130:[email protected] myself,master - 0 1615365162000 1 connected 0-5460 1379d2b20f26ab13d53068d276ec5d988b7a0273 100.64.122.197:[email protected] slave 1895cb4b9c31b848666c61000e502f55a29a8255 0 1615365163000 2 connected e83446c368839c5fdccf5f70e3b1004eb67cb651 100.107.55.68:[email protected] master - 0 1615365164014 3 connected 10923-16383

以上是 kubernetes redis-cluster生产实践笔记的全部内容，来源链接： utcz.com/a/121134.html