1、环境
主机列表：
node1 10.10.10.25
node2 10.10.10.26
node3 10.10.10.27
node1 k8s的管理节点 nfs服务器
node2 k8s的工作节点
node3 k8s的工作节点
2、创建数据目录
在node1基于nfs创建mongo数据目录：
mkdir /k8s/
chmod 777 k8s
cd k8s
mkdir /k8s/mongo-1
mkdir /k8s/mongo-1/data
mkdir /k8s/mongo-1/key
mkdir /k8s/mongo-2
mkdir /k8s/mongo-2/data
mkdir /k8s/mongo-2/key
vi /etc/exports
/k8s/ *(insecure,rw,sync,no_root_squash)
systemctl restart nfs
3、检查nfs目录是否创建成功
登陆node2 或者node3执行以下命令
showmount -e 10.10.10.25
可以看到以下结果
image.png
4、编写ConfigMap配置文件
mongod.conf 配置
apiVersion: v1
data:
mongod.conf: |-
dbpath=/mongo/data
#logpath=/mongo/log/mongodb.log
pidfilepath=/mongo/key/master.pid
directoryperdb=true
logappend=true
bind_ip=0.0.0.0
port=27017
kind: ConfigMap
metadata:
name: mongodb-conf
namespace: default
5、配置pv
mongo-1挂载卷配置
apiVersion: v1
kind: PersistentVolume
metadata:
name: mongo-pv-1
spec:
accessModes:
- ReadWriteOnce
capacity:
storage: 3Gi
nfs:
path: /k8s/mongo-1
readOnly: false
server: 10.10.10.25
mongo-2挂载卷配置
apiVersion: v1
kind: PersistentVolume
metadata:
name: mongo-pv-2
spec:
accessModes:
- ReadWriteOnce
capacity:
storage: 3Gi
nfs:
path: /k8s/mongo-2
readOnly: false
server: 10.10.10.25
6、创建有状态副本集StatefulSet
mongo-1有状态副本集
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
namespace: default
name: mongo-1
app: mongo-cluster
spec:
selector:
matchLabels:
name: mongo-1
serviceName: "mongo-1"
replicas: 1
podManagementPolicy: Parallel
template:
metadata:
labels:
name: mongo-1
app: mongo-cluster
spec:
terminationGracePeriodSeconds: 10
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: "name"
operator: In
values:
- mongo
topologyKey: "kubernetes.io/hostname"
containers:
- name: mongo
image: mongo:4.2
imagePullPolicy: IfNotPresent
command:
- mongod
- "-f"
- "/etc/mongod.conf"
- "--bind_ip_all"
- "--replSet"
- rs0
ports:
- containerPort: 27017
volumeMounts:
- name: mongo-cnf-volume
mountPath: /etc/mongod.conf/
subPath: mongod.conf
- name: mongo-dir
mountPath: /mongo
volumes:
- name: mongo-cnf-volume #映射configMap信息
configMap:
name: mongodb-conf
items:
- key: mongod.conf
path: mongod.conf
volumeClaimTemplates:
- metadata:
name: mongo-dir
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 3Gi
mongo-2有状态副本集
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
namespace: default
name: mongo-2
app: mongo-cluster
spec:
selector:
matchLabels:
name: mongo-2
serviceName: "mongo-2"
replicas: 1
podManagementPolicy: Parallel
template:
metadata:
labels:
name: mongo-2
app: mongo-cluster
spec:
terminationGracePeriodSeconds: 10
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: "name"
operator: In
values:
- mongo
topologyKey: "kubernetes.io/hostname"
containers:
- name: mongo
image: mongo:4.2
imagePullPolicy: IfNotPresent
command:
- mongod
- "-f"
- "/etc/mongod.conf"
- "--bind_ip_all"
- "--replSet"
- rs0
ports:
- containerPort: 27017
volumeMounts:
- name: mongo-cnf-volume
mountPath: /etc/mongod.conf/
subPath: mongod.conf
- name: mongo-dir
mountPath: /mongo
volumes:
- name: mongo-cnf-volume #映射configMap信息
configMap:
name: mongodb-conf
items:
- key: mongod.conf
path: mongod.conf
volumeClaimTemplates:
- metadata:
name: mongo-dir
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 3Gi
7、创建无头服务HeadlessService
apiVersion: v1
kind: Service
metadata:
namespace: default
name: mongo-1 #需要与podname同名
labels:
name: mongo
spec:
ports:
- name: mongo-port
port: 27017
clusterIP: None
selector:
name: mongo-1
apiVersion: v1
kind: Service
metadata:
namespace: default
name: mongo-2 #需要与podname同名
labels:
name: mongo
spec:
ports:
- name: mongo-port
port: 27017
clusterIP: None
selector:
name: mongo-2
8、创建一个busybox，执行以下命令查看dns信息
apiVersion: v1
kind: Pod
metadata:
name: busybox
namespace: default
spec:
containers:
- name: busybox
image: busybox:1.32
command:
- sleep
- "3600"
imagePullPolicy: IfNotPresent
restartPolicy: Always
进入busybox容器内，执行以下命令
nslookup mongo-1
image.png
nslookup mongo-2
image.png
9、初始化集群
进入mongo-1内部、执行一下命令：
#连接mongo数据库
mongo
#初始化集群
rs.initiate({ _id:"rs0", // replSet指定的名称
members:[{ _id:0, host:"mongo-1.default.svc.cluster.local:27017" // 主节点ip与端口,
}]
})
image.png
#mongo-2加入集群
rs.add("mongo-2.default.svc.cluster.local:27017")// 将mongo-2加入集群
image.png
ok为1为成功。
10、创建外部访问的NodePort Service
master NodePort Service
apiVersion: v1
kind: Service
metadata:
name: mongo-1-front-service
labels:
name: mongo-1
namespace: default
spec:
selector:
name: mongo-1
type: NodePort
externalTrafficPolicy: Cluster
ports:
- name: mongo-http
nodePort: 30882
port: 27017
protocol: TCP
targetPort: 27017
slave NodePort Service
apiVersion: v1
kind: Service
metadata:
name: mongo-2-front-service
labels:
name: mongo-2
namespace: default
spec:
selector:
name: mongo-2
type: NodePort
externalTrafficPolicy: Cluster
ports:
- name: mongo-http
nodePort: 30883
port: 27017
protocol: TCP
targetPort: 27017
此时，mongo集群的部署已经完成，可以使用外部客户端访问了。
