搭建Docker Swarm集群

一.环境准备
1.3台服务器，配置静态ip，主机名，绑定hosts
hostnamectl set-hostname <主机名>
1.1.1.3 node1
1.1.1.4 node2
1.1.1.5 node3

2.关闭selinux，iptables，firewalld

# iptables -F
# systemctl disable firewalld
# systemctl stop firewalld
# vim /etc/sysconfig/selinux
SELINUX=disabled

二.安装Docker
安装必要的工具

yum install -y yum-utils device-mapper-persistent-data lvm2

添加Docker的存储库

yum-config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo

安装

yum install -y docker-ce

配置国内镜像源
在/etc/docker/目录下新建文件daemon.json
配置内容如下

{
 "registry-mirrors": ["https://registry.docker-cn.com"]
}

重新载入配置并启动Docker

systemctl daemon-reload
systemctl start docker

三.开始搭建Swarm集群
1.初始化集群，执行初始化的服务器默认为管理节点

docker swarm init --advertise-addr 1.1.1.3
Swarm initialized: current node (l9mveonpubd7dugdpbrraq1q3) is now a manager.

To add a worker to this swarm, run the following command:

    docker swarm join \
	--token SWMTKN-1-6co0gt73819kggxcsnqfr7j32ner9tdl7wi3mf0dnvu0o2bknz-ablwzg4iahibpq6r9dquwppvr \
	1.1.1.3:2377         --执行该条命令，以worker的身份加入集群

To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.

查看Swarm集群激活状态

docker info | grep -i swarm
Swarm: active

如果出现以下提示

WARNING: IPv4 forwarding is disabled
WARNING: bridge-nf-call-iptables is disabled
WARNING: bridge-nf-call-ip6tables is disabled
Swarm: active

解决方法如下
修改/etc/sysctl.conf文件,内容如下

net.ipv4.ip_forward = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
net.bridge.bridge-nf-call-arptables = 1

执行sysctl -p，生效

2.将其他服务器加入到集群
node2以管理者身份加入集群

docker swarm join \
    --token SWMTKN-1-6co0gt73819kggxcsnqfr7j32ner9tdl7wi3mf0dnvu0o2bknz-cy4uh1dfa4l4s3cng0yg5lujn \
    1.1.1.3:2377

node3以工作者身份加入集群

docker swarm join \
	--token SWMTKN-1-6co0gt73819kggxcsnqfr7j32ner9tdl7wi3mf0dnvu0o2bknz-ablwzg4iahibpq6r9dquwppvr \
	1.1.1.3:2377

查看集群成员

# docker node ls
ID                           HOSTNAME  STATUS  AVAILABILITY  MANAGER STATUS
5bq2l28mb6kdl88t4lzjdz6xy    node2     Ready   Active        Reachable
l0j2psjlfs3hoywwjzlk785c2    node3     Ready   Active        
l9mveonpubd7dugdpbrraq1q3 *  node1     Ready   Active        Leader

新建一个overlay网络
Swarm集群默认的ingress网络不支持服务发现，所以重新创建一个网络。

# docker network create --attachable --driver overlay microservice
rjxxb5gz6gbx2bc50077k9uby

查看集群所有网络

# docker network ls
NETWORK ID          NAME                DRIVER              SCOPE
0aa88eeb09b2        bridge              bridge              local
38c8f825c2a8        docker_gwbridge     bridge              local
2df53c0f2885        host                host                local
sce8abmnyz7j        ingress             overlay             swarm
rjxxb5gz6gbx        microservice        overlay             swarm
2e8597069347        none                null                local

Docker Swarm集群搭建完成

四.搭建consul集群和server高可用
官方建议在运行Docker容器时，用host网络模式
1.在node1上执行，安装consul server

docker run -d -p 8300:8300 -p 8301:8301/tcp -p 8302:8302/tcp -p 8301:8301/udp \
-p 8302:8302/udp -p 8500:8500 -p 8600:8600/udp -p 8600:8600/tcp \
--restart=always \
-h node1 \
--name consul  \
--network microservice \
-v /data/consul:/consul/data \
-v /etc/localtime:/etc/localtime:ro \
consul agent \
-server \
-bootstrap-expect=2 \
-node=node1 \
-rejoin \
-client 0.0.0.0 \
-advertise 1.1.1.3 \
-ui

2.在node2上执行，安装consul server，与上面的命令有区别，没有-ui选项

docker run -d -p 8300:8300 -p 8301:8301/tcp -p 8302:8302/tcp -p 8301:8301/udp \
-p 8302:8302/udp -p 8500:8500 -p 8600:8600/udp -p 8600:8600/tcp \
--restart=always \
-h node2 \
--name consul  \
--network microservice \
-v /etc/localtime:/etc/localtime:ro \
consul agent \
-server \
-node=node2 \
-rejoin \
-client 0.0.0.0 \
-join 1.1.1.3 \
-advertise 1.1.1.4

3.在node3上执行，安装consul client

docker run -d -p 8300:8300 -p 8301:8301/tcp -p 8302:8302/tcp -p 8301:8301/udp \
-p 8302:8302/udp -p 8500:8500 -p 8600:8600/udp -p 8600:8600/tcp \
--restart=always \
-h node3 \
--name consul  \
--network microservice \
-v /etc/localtime:/etc/localtime:ro \
consul agent \
-node=node3 \
-rejoin \
-client 0.0.0.0 \
-join 1.1.1.3 \
-advertise 1.1.1.5

4.在node1上查看consul集群节点信息

# docker exec consul consul members
Node   Address       Status  Type    Build  Protocol  DC   Segment
node1  1.1.1.3:8301  alive   server  1.4.4  2         dc1  <all>
node2  1.1.1.4:8301  alive   server  1.4.4  2         dc1  <all>
node3  1.1.1.5:8301  alive   client  1.4.4  2         dc1  <default>

五.搭建服务注册，采用registrator实现
在node1,node2,node3按顺序执行以下命令

docker run -d \
    --restart=always \
    --name=registrator \
    --net=host \
    --volume=/var/run/docker.sock:/tmp/docker.sock \
    -v /etc/localtime:/etc/localtime:ro \
    doubleshit/registrator:v7 \
    -resync 60 \
    -cleanup \
    -internal \
    -ip 1.1.1.3 \
    -useIpFromNetworkName=microservice \
      consul://1.1.1.3:8500

可以通过访问：http://1.1.1.3:8500 入到consul的ui界面查看运行的服务或者通过curl http://1.1.1.3:8500/v1/catalog/services在终端查看

查看详细的服务信息

# curl http://1.1.1.3:8500/v1/catalog/service/<服务名>

集群搭建完成