docker swarm 搭建及跨主机网络互连案例分析

时间:2022-07-11 09:08:14

准备工作

安装docker,不建议直接使用Docker官方的yum install docker

wget http://yum.dockerproject.org/repo/main/centos/7/Packages/docker-engine-selinux-1.12.6-1.el7.centos.noarch.rpm
wget http://yum.dockerproject.org/repo/main/centos/7/Packages/docker-engine-1.12.6-1.el7.centos.x86_64.rpm
wget http://yum.dockerproject.org/repo/main/centos/7/Packages/docker-engine-debuginfo-1.12.6-1.el7.centos.x86_64.rpm
yum localinstall -y docker-engine-selinux-1.12.-.el7.centos.noarch.rpm docker-engine-1.12.-.el7.centos.x86_64.rpm docker-engine-debuginfo-1.12.-.el7.centos.x86_64.rpm

http://download.csdn.net/detail/yiyu1/9875469

一、开放相关端口

firewall-cmd --zone=public --add-port=/tcp --permanent
firewall-cmd --zone=public --add-port=/tcp --permanent
firewall-cmd --zone=public --add-port=/udp --permanent
firewall-cmd --zone=public --add-port=/tcp --permanent
firewall-cmd --zone=public --add-port=/udp --permanent
firewall-cmd --reload

二、修改配置 版本不同,可以忽略

vi /etc/docker/daemon.json
"live-restore": false

有关集群的docker命令如下:

  1. docker swarm:集群管理,子命令有init, join,join-token, leave, update
  2. docker node:节点管理,子命令有demote, inspect,ls, promote, rm, ps, update
  3. docker service:服务管理,子命令有create, inspect, ps, ls ,rm , scale, update
  4. docker stack/deploy:试验特性,用于多应用部署

创建swarm 集群

查看docker swarm 命令说明

# docker swarm -h
Flag shorthand -h has been deprecated, please use --help
Usage: docker swarm COMMAND
Manage Docker Swarm
Options:
--help Print usage
Commands:
init Initialize a swarm
join Join a swarm as a node and/or manager
join-token Manage join tokens
update Update the swarm
leave Leave a swarm
Run 'docker swarm COMMAND --help' for more information on a command.

创建集群

#命令格式: docker swarm init --listen-addr <MANAGER-IP>:<PORT>
[root@centos-web ~]# docker swarm init --listen-addr 172.18.30.29:
Swarm initialized: current node (a60d5c3ttymvtozr46uvk17q4) is now a manager.
docker swarm init --advertise-addr 122.62.24.161 --listen-addr 122.62.24.161:2377

查看集群

# docker node ls
ID HOSTNAME MEMBERSHIP STATUS AVAILABILITY MANAGER STATUS
a60d5c3ttymvtozr46uvk17q4 * centos-web Accepted Ready Active Leader

获取加入集群命令token

#docker swarm join-token manager
docker swarm join \
--token ********* \
172.18.30.29:

子节点加入集群

docker swarm join \
--token ********* \
172.18.30.29:

删除集群子节点

docker node demote ***
docker swarm leave --force(节点上)
docker node rm --force(manager上)

创建一个overlay 跨主机网络

查看当前网络

[root@H-T- ~]# docker network ls
NETWORK ID NAME DRIVER SCOPE
47d05fd82e77 bridge bridge local
76e0302f7929 docker_gwbridge bridge local
81c22e7f1b48 host host local
4enpk54t9oy2 ingress overlay swarm
c80502ba134b none null local

创建一个新的overlay网络

#swarm上默认已有一个名为ingress的overlay 网络

 [root@centos-node4 ~]# docker network create --driver overlay docker-net
2o223p435s5glx69dw211u5d2
[root@centos-web ~]# docker network ls
NETWORK ID NAME DRIVER SCOPE
abec77415f48 bridge bridge local
aoqs3p835s5g docker-net overlay swarm
e2fff9d572a6 docker_gwbridge bridge local
166bd71f7d0e host host local
9gr6bfff1rv9 ingress overlay swarm
1d2bfc590294 none null local

docker-net就是新创建的网络

在新的跨主机overlay 网络(docker-net)上创建应用

部署应用

docker service create --constraint=node.HOSTNAME==H-C-42 --replicas  --publish : --name lvs --network=docker-net inits/lvs

指定一个节点部署应用

# docker node ls
ID HOSTNAME MEMBERSHIP STATUS AVAILABILITY MANAGER STATUS
a60d5c3ttymvtozr46uvk17q4 * centos-web Accepted Ready Active Leader
#docker service create --constraint=node.HOSTNAME==centos-web --replicas 1 --publish 28152:22 --publish 8152:8080 --name web --network=docker-net  inits/source2.0.3
ce2qc5paxmfa2gxrtu320v5m6

–replicas 副本数量
–publish 服务发现,端口映射
–mount 挂载文件

--mount type=bind,src=/home,dst=/home,readonly
–name service名称

#前提节点服务器上images上有inits/lvs镜像

部署结果

docker service ls
ID NAME REPLICAS IMAGE COMMAND
5lgdq3ihiez0 lvs / inits/lvs
[root@centos-node4 ~]# docker service tasks helloworld
ID NAME SERVICE IMAGE LAST STATE DESIRED STATE NODE
eul3bus45qz3b555wekotdmo5 lvs. helloworld alpine Running seconds Running centos-node5
55uhq6xxcv53xlkqv2f0be9b9 lvs. helloworld alpine Running seconds Running centos-node4

分别在运行两个节点上查看容器运行情况

#ssh 172.18.30.29
[root@H-T- ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
389f9eefe455 inits/lvs:latest "/usr/sbin/sshd -D" minutes ago Up minutes /tcp lvs..aco0uuf1of90tl31skh6a6ndv
#ssh 172.18.30.12
[root@H-C- ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
af573b0ae945 inits/lvs:latest "/usr/sbin/sshd -D" minutes ago Up minutes /tcp lvs..0a1s4la7prc3qmfdfdqazitsn

其他

docker service ls 查看集群列表
docker service ps lvs 查看集群下所有节点状态
docker service rm lvs 删除集群
docker service inspect --pretty lvs 集群属性
docker service scale lvs=4 #扩容集群节点数量

测试两个主机的网络是否能互通

# docker exec -ti lvs..0a1s4la7prc3qmfdfdqazitsn sh
/ # ping lvs..aco0uuf1of90tl31skh6a6ndv
PING lvs..aco0uuf1of90tl31skh6a6ndv (10.0.9.3): data bytes
bytes from 10.0.9.3: seq= ttl= time=0.514 ms
bytes from 10.0.9.3: seq= ttl= time=0.508 ms
bytes from 10.0.9.3: seq= ttl= time=0.381 ms
bytes from 10.0.9.3: seq= ttl= time=0.408 ms
^C
--- lvs..aco0uuf1of90tl31skh6a6ndv ping statistics ---
packets transmitted, packets received, % packet loss
round-trip min/avg/max = 0.381/0.452/0.514 ms

同理去另一台主机测试网络是否正常

或已暴露22端口,直接

ssh 172.18.30.29 -p
ifconfig
172.18..29内部ip
ping 172.18..12内部ip
ssh 172.18.30.12 -p
ifconfig
172.18..12内部ip
ping 172.18..29内部ip

dokcer swarm自带的负载均衡

创建一组服务

docker service create --replicas  --name nginx -p :80 --network=docker-net inits/nginx

测试服务

浏览器访问:
http://172.18.30.29
刷新浏览器
系统将轮询访问集群节点下服务