kubesphere+k8s+helm安装RabbitMQ集群+镜像模式

1.安装k8s多节点集群#

1.下载Rabbitmq的chart包#

添加bitnamid仓库

1
helm repo add bitnami https://charts.bitnami.com/bitnami

查询chart
Terminal window
```
1
helm search repo bitnami
```
创建rabbitmq的工作目录
Terminal window
```
1
mkdir -p /opt/rabbitmq
```
使用helm拉取rabbitmq
Terminal window
```
1
helm pull bitnami/rabbitmq
```

解压

1
tar -xvf 拉取的文件名-版本变更会导致文件名变化

2.编辑配置文件#

设置管理员密码

配置文件方式

1
auth:
2
  ## @param auth.username RabbitMQ application username
3
  ## ref: https://github.com/bitnami/bitnami-docker-rabbitmq#environment-variables
4
  ##
5
  username: admin
6

7
  ## @param auth.password RabbitMQ application password
8
  ## ref: https://github.com/bitnami/bitnami-docker-rabbitmq#environment-variables
9
  ##
10
  password: "77589910"
11
  ## @param auth.existingPasswordSecret Existing secret with RabbitMQ credentials (must contain a value for `rabbitmq-password` key)
12
  ## e.g:
13
  ## existingPasswordSecret: name-of-existing-secret
14
  ##
15
  existingPasswordSecret: ""
16

17
  ## @param auth.erlangCookie Erlang cookie to determine whether different nodes are allowed to communicate with each other
18
  ## ref: https://github.com/bitnami/bitnami-docker-rabbitmq#environment-variables
19
  ##
20
  erlangCookie: "secretcookie123"
21
  ## @param auth.existingErlangSecret Existing secret with RabbitMQ Erlang cookie (must contain a value for `rabbitmq-erlang-cookie` key)
22
  ## e.g:
23
  ## existingErlangSecret: name-of-existing-secret
24
  ##
25
  existingErlangSecret: ""

安装时set方式指定

1
--set auth.username=admin,auth.password=77589910,auth.erlangCookie=secretcookie123

设置启动副本数
Terminal window
```
1
replicaCount: 3
```
image-20220123204103644

设置持久化存储

1
persistence:
2
  ## @param persistence.enabled Enable RabbitMQ data persistence using PVC
3
  ##
4
  enabled: true
5

6
  ## @param persistence.storageClass PVC Storage Class for RabbitMQ data volume
7
  ## If defined, storageClassName: <storageClass>
8
  ## If set to "-", storageClassName: "", which disables dynamic provisioning
9
  ## If undefined (the default) or set to null, no storageClassName spec is
10
  ##   set, choosing the default provisioner.  (gp2 on AWS, standard on
11
  ##   GKE, AWS & OpenStack)
12
  ##
13
  storageClass: "local"
14
  ## @param persistence.selector Selector to match an existing Persistent Volume
15
  ## selector:
16
  ##   matchLabels:
17
  ##     app: my-app
18
  ##
19
  selector: {}
20
  ## @param persistence.accessMode PVC Access Mode for RabbitMQ data volume
21
  ##
22
  accessMode: ReadWriteOnce
23

24
  ## @param persistence.existingClaim Provide an existing PersistentVolumeClaims
25
  ## The value is evaluated as a template
26
  ## So, for example, the name can depend on .Release or .Chart
27
  ##
28
  existingClaim: ""
29

30
  ## @param persistence.size PVC Storage Request for RabbitMQ data volume
31
  ## If you change this value, you might have to adjust `rabbitmq.diskFreeLimit` as well
32
  ##
33
  size: 60Gi
34

35
  ## @param persistence.volumes Additional volumes without creating PVC
36
  ##  - name: volume_name
37
  ##    emptyDir: {}
38
  ##
39
  volumes: []
40
  ## @param persistence.annotations Persistence annotations. Evaluated as a template
41
  ## Example:
42
  ## annotations:
43
  ##   example.io/disk-volume-type: SSD
44
  ##
45
  annotations: {}

设置集群访问方式
```
1
service:
2
  ## @param service.type Kubernetes Service type
3
  ##
4
  type: NodePort
```
在实操中我发现,如果service.type设置成ClusterIP,等待集群启动完成后,再手动添加NodePort的访问方式,手动删除一个容器后,容器在自动恢复时会报错,然后就是不停的重启

设置集群意外宕机强制重启

当rabbitmq启用持久化存储时，若rabbitmq所有pod同时宕机，将无法重新启动，因此有必要提前开启clustering.forceBoot

1
clustering:
2
  ## @param clustering.enabled Enable RabbitMQ clustering
3
  ##
4
  enabled: true
5
  ## @param clustering.addressType Switch clustering mode. Either `ip` or `hostname`
6
  ##
7
  addressType: hostname
8
  ## @param clustering.rebalance Rebalance master for queues in cluster when new replica is created
9
  ## ref: https://www.rabbitmq.com/rabbitmq-queues.8.html#rebalance
10
  ##
11
  rebalance: false
12

13
  ## @param clustering.forceBoot Force boot of an unexpectedly shut down cluster (in an unexpected order).
14
  ## forceBoot executes 'rabbitmqctl force_boot' to force boot cluster shut down unexpectedly in an unknown order
15
  ## ref: https://www.rabbitmq.com/rabbitmqctl.8.html#force_boot
16
  ##
17
  forceBoot: true
18

19
  ## @param clustering.partitionHandling Switch Partition Handling Strategy. Either `autoheal` or `pause-minority` or `pause-if-all-down` or `ignore`
20
  ## ref: https://www.rabbitmq.com/partitions.html#automatic-handling
21
  ##
22
  partitionHandling: autoheal

指定时区

1
extraEnvVars:
2
  - name: TZ
3
    value: "Asia/Shanghai"

3.部署RabbitMQ#

在kubesphere后台创建命名空间demo-project

使用kubectl命令创建的在kubesphere后台可能会不显示
安装
- 使用配置文件中配置的管理员账号密码
  Terminal window
```
1
helm install rabbitmq -n demo-project .
```
  helm install后的rabbitmq只是一个名称,起什么名称都可以
  
  命令最后有一个点
- 通过set指定管理员账号密码—后期upgrade时也须指定上参数
  Terminal window
```
1
helm install rabbitmq -n demo-project . \
2
--set auth.username=admin,auth.password=77589910,auth.erlangCookie=secretcookie123
```
image-20220123205912776
查看安装状态
- 查看安装进度
  Terminal window
```
1
kubectl get pod -n demo-project -w
```
  image-20220123210139110
- 查看svc
  Terminal window
```
1
kubectl get svc -n demo-project
```
- 查看集群状态
  - 进入pod
    Terminal window
    1 kubectl exec -it -n demo-project rabbitmq-0 -- bash
  - 查看集群状态
    Terminal window
    1 rabbitmqctl cluster_status
  - 查看当前策略
    Terminal window
    1 rabbitmqctl list_policies
  - 设置集群名称
    Terminal window
    1 rabbitmqctl set_cluster_name [cluster_name]

4.配置镜像模式实现高可用#

使用命令

1
rabbitmqctl set_policy ha-all "^" '{"ha-mode":"all" , "ha-sync-mode":"automatic"}'

使用控制台

image-20220121183803914

5.清理RabbitMQ集群#

卸载

1
helm uninstall rabbitmq -n demo-project

删除pvc

1
kubectl delete pvc -n demo-project data-rabbitmq-0 data-rabbitmq-1 data-rabbitmq-2

6.参考文献#

https://blog.csdn.net/qq_14999375/article/details/119085363

音乐

音乐

1.安装k8s多节点集群#

1.下载Rabbitmq的chart包#

2.编辑配置文件#

3.部署RabbitMQ#

4.配置镜像模式实现高可用#

5.清理RabbitMQ集群#

6.参考文献#

文章分享