环境准备
172.31.0.201 compo-node1.local
172.31.0.202 compo-node2.local
172.31.0.202 compo-node2.local
注意：docker-swarm是通过主机名区分，所以需要改主机名
[root@long-ubuntu ~]# hostnamectl set-hostname compo-node1.local
[root@long-ubuntu ~]# hostnamectl set-hostname compo-node2.local
[root@long-ubuntu ~]# hostnamectl set-hostname compo-node3.local
改好主机名后，然后在某个节点做初始化
[root@compo-node1 ~]# docker swarm init --advertise-addr 172.31.0.201
Swarm initialized: current node (pjpobxje1z5h2v1ym0lwmszjf) is now a manager.
To add a worker to this swarm, run the following command:
docker swarm join --token SWMTKN-1-4ksr4at7jw6op59marbqorzez47vzl6gdbq3j2iehqvyobk3yw-3eks198rxc2dwh3ypw0mwniyd 172.31.0.201:2377
To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.
列出所有节点
[root@compo-node1 ~]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
pjpobxje1z5h2v1ym0lwmszjf * compo-node1.local Ready Active Leader 19.03.15
添加节点
[root@compo-node2 ~]# docker swarm join --token SWMTKN-1-4ksr4at7jw6op59marbqorzez47vzl6gdbq3j2iehqvyobk3yw-3eks198rxc2dwh3ypw0mwniyd 172.31.0.201:2377
This node joined a swarm as a worker.
[root@compo-node3 ~]# docker swarm join --token SWMTKN-1-4ksr4at7jw6op59marbqorzez47vzl6gdbq3j2iehqvyobk3yw-3eks198rxc2dwh3ypw0mwniyd 172.31.0.201:2377
This node joined a swarm as a worker.
添加label
# 先查看所有的节点
[root@compo-node1 ~]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
pjpobxje1z5h2v1ym0lwmszjf * compo-node1.local Ready Active Leader 19.03.15
pjk55jhzcg50jyyvpofyknsoh compo-node2.local Ready Active 19.03.15
m8twd2rx7slhmh3hvwfcufxv9 compo-node3.local Ready Active 19.03.15
# 添加
[root@compo-node1 ~]# docker node update --label-add name=compo-node1 compo-node1.local
compo-node1.local
[root@compo-node1 ~]# docker node update --label-add name=compo-node2 compo-node2.local
compo-node2.local
[root@compo-node1 ~]# docker node update --label-add name=compo-node3 compo-node3.local
compo-node3.local
将其他节点提升为manager角色以实现高可用
[root@compo-node1 ~]# docker node promote compo-node2.local
Node compo-node2.local promoted to a manager in the swarm.
[root@compo-node1 ~]# docker node promote compo-node3.local
Node compo-node3.local promoted to a manager in the swarm.
再次查看所有节点
[root@compo-node1 ~]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
pjpobxje1z5h2v1ym0lwmszjf * compo-node1.local Ready Active Leader 19.03.15
pjk55jhzcg50jyyvpofyknsoh compo-node2.local Ready Active Reachable 19.03.15
m8twd2rx7slhmh3hvwfcufxv9 compo-node3.local Ready Active Reachable 19.03.15
查看node信息
[root@compo-node1 ~]# docker node inspect compo-node2.local
创建网络
# 帮助
[root@compo-node1 ~]# docker network --help
# 创建网络的命令
[root@compo-node1 ~]# docker network create -d overlay --subnet=10.200.0.0/21 --gateway=10.200.0.1 --attachable long-net
mqqmq52rjeiv970dkbgbq5rhl
查看网络
[root@compo-node3 ~]# docker network ls
NETWORK ID NAME DRIVER SCOPE
fda09d67e839 bridge bridge local
7dba8f8f069e docker_gwbridge bridge local
663fea23b067 host host local
o7hubsieqobu ingress overlay swarm
mqqmq52rjeiv long-net overlay swarm
e3dd1c687900 none null local
验证网络
[root@compo-node1 ~]# docker network inspect long-net
[
{
"Name": "long-net",
"Id": "mqqmq52rjeiv970dkbgbq5rhl",
"Created": "2021-07-22T07:33:45.229235855Z",
"Scope": "swarm",
"Driver": "overlay",
"EnableIPv6": false,
"IPAM": {
"Driver": "default",
"Options": null,
"Config": [
{
"Subnet": "10.200.0.0/21",
"Gateway": "10.200.0.1"
}
]
},
"Internal": false,
"Attachable": true,
"Ingress": false,
"ConfigFrom": {
"Network": ""
},
"ConfigOnly": false,
"Containers": null,
"Options": {
"com.docker.network.driver.overlay.vxlanid_list": "4097"
},
"Labels": null
}
]
创建容器测试
[root@compo-node1 ~]# docker service create --replicas 3 -p 8888:80 --network long-net --name nginx nginx:1.18-alpine
image nginx:1.18-alpine could not be accessed on a registry to record
its digest. Each node will access nginx:1.18-alpine independently,
possibly leading to different nodes running different
versions of the image.
p45rj6a1pb76vrlmsarweyexx
overall progress: 3 out of 3 tasks
1/3: running [==================================================>]
2/3: running [==================================================>]
3/3: running [==================================================>]
verify: Service converged
对上面的命令参数做下解释
--replicas 3表示要启动3个服务
--name nginx表示服务的名称为nginx，
nginx:1.18-alpine表示根据这个镜像创建的容器(服务)
验证
[root@compo-node1 ~]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
p45rj6a1pb76 nginx replicated 3/3 nginx:1.18-alpine *:8888->80/tcp
验证端口监听，每个服务器都会监听service端口8888
[root@compo-node1 ~]# ss -tanl
LISTEN 0 128 [::]:111 [::]:*
LISTEN 0 128 [::]:22 [::]:*
LISTEN 0 128 [::]:35927 [::]:*
LISTEN 0 128 *:8888 *:*
访问测试
[root@compo-node1 ~]# curl 172.31.0.201:8888
验证service
[root@compo-node1 ~]# docker service ps p45rj6a1pb76
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
bdpx7xicb6ui nginx.1 nginx:1.18-alpine compo-node2.local Running Running 7 minutes ago
wi3gawal92t5 nginx.2 nginx:1.18-alpine compo-node3.local Running Running 7 minutes ago
nhg6bg1b1tnj nginx.3 nginx:1.18-alpine compo-node2.local Running Running 5 minutes ago
验证高可用：将容器所在的服务器或docker关闭，验证pod副本高可用
# 停止
[root@compo-node1 ~]# systemctl stop docker
# 查看
[root@compo-node2 ~]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
pjpobxje1z5h2v1ym0lwmszjf compo-node1.local Down Active Unreachable 19.03.15
pjk55jhzcg50jyyvpofyknsoh * compo-node2.local Ready Active Leader 19.03.15
m8twd2rx7slhmh3hvwfcufxv9 compo-node3.local Ready Active Reachable 19.03.15
启动
[root@compo-node1 ~]# systemctl start docker
# 查看
[root@compo-node1 ~]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
pjpobxje1z5h2v1ym0lwmszjf * compo-node1.local Ready Active Reachable 19.03.15
pjk55jhzcg50jyyvpofyknsoh compo-node2.local Ready Active Leader 19.03.15
m8twd2rx7slhmh3hvwfcufxv9 compo-node3.local Ready Active Reachable 19.03.15
删除（慎用，随便一台集群机器删除就所有机器都没有了）
[root@compo-node1 ~]# docker service rm nginx
nginx
再次查看
[root@compo-node1 ~]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
tmm1faqhlt7j nginx-web replicated 10/10 nginx:1.18-alpine *:8880->80/tcp
[root@compo-node1 ~]# docker service rm nginx-web
nginx-web
[root@compo-node1 ~]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
[root@compo-node2 ~]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
[root@compo-node3 ~]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
范例：创建一个10个应用的
[root@compo-node1 ~]# docker service create --replicas 10 -p 8888:80 --network long-net --name nginx nginx:1.18-alpine
image nginx:1.18-alpine could not be accessed on a registry to record
its digest. Each node will access nginx:1.18-alpine independently,
possibly leading to different nodes running different
versions of the image.
ljsbgphte9wr4oj6f3lvld8vh
overall progress: 10 out of 10 tasks
1/10: running [==================================================>]
2/10: running [==================================================>]
3/10: running [==================================================>]
4/10: running [==================================================>]
5/10: running [==================================================>]
6/10: running [==================================================>]
7/10: running [==================================================>]
8/10: running [==================================================>]
9/10: running [==================================================>]
10/10: running [==================================================>]
verify: Service converged
查看
[root@compo-node1 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
[root@compo-node2 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
da7f30c5f7f3 nginx:1.18-alpine "/docker-entrypoint.…" 10 minutes ago Up 10 minutes 80/tcp nginx.4.rh4j5dxi5mb5e9a8sdnufo405
004e071d1288 nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.1.25zkndu23ea16hylehzdj9h4m
395219dcd109 nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.9.tb5p9u6dlfebj7bqph3vlqoms
2bccdaea3cd8 nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.3.ibvdx58d8hrqqqhi78omlgrp2
df741faa77b4 nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.5.fbmapne5e78ky0faszhpsevaa
[root@compo-node3 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
8bfc03216f76 nginx:1.18-alpine "/docker-entrypoint.…" 10 minutes ago Up 10 minutes 80/tcp nginx.8.eoyokyiu4avrzoyh2hx7w49vb
5d86a9864778 nginx:1.18-alpine "/docker-entrypoint.…" 10 minutes ago Up 10 minutes 80/tcp nginx.10.pb1hfo1d05d6xq77r0iq6u45b
6998be3d0dfe nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.7.q93goqhwa2cxm96k51simv2xi
5c8720961d0b nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.2.tmwt38hhao5w9k5r3cxmzrgsw
a9b3af2e4beb nginx:1.18-alpine "/docker-entrypoint.…" 11 minutes ago Up 11 minutes 80/tcp nginx.6.ux6uqk21zl0ssnxwjncmj8n2y
伸缩
[root@compo-node3 ~]# docker service scale nginx=3
nginx scaled to 3
overall progress: 3 out of 3 tasks
1/3: running [==================================================>]
2/3: No such image: nginx:1.18-alpine
3/3: running [==================================================>]
verify: Service converged
再次查看
[root@compo-node1 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
[root@compo-node2 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
004e071d1288 nginx:1.18-alpine "/docker-entrypoint.…" 13 minutes ago Up 13 minutes 80/tcp nginx.1.25zkndu23ea16hylehzdj9h4m
2bccdaea3cd8 nginx:1.18-alpine "/docker-entrypoint.…" 14 minutes ago Up 13 minutes 80/tcp nginx.3.ibvdx58d8hrqqqhi78omlgrp2
[root@compo-node3 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
5c8720961d0b nginx:1.18-alpine "/docker-entrypoint.…" 13 minutes ago Up 13 minutes 80/tcp nginx.2.tmwt38hhao5w9k5r3cxmzrgsw
总结
有个需求，假如nginx容器启用有300个，此时需要缩容，就得先缩容到100，然后缩容到20个，而不是300个容器上来就缩容3个，这样只会减少3个，而不是只剩3个，跟创建容器个数相反
[root@compo-node1 ~]# docker service scale nginx=3
nginx scaled to 3
overall progress: 3 out of 3 tasks
1/3: starting container failed: Address already in use
2/3:
3/3:
verify: Service converged
[root@compo-node1 ~]# docker service scale nginx=100
nginx scaled to 100
overall progress: 100 out of 100 tasks
verify: Service converged