部署Swarm Mode集群

环境准备

主机名（角色）	IP
swarm-manager	172.16.100.20
swarm-node1	172.16.100.22
swarm-node2	172.16.100.22

加入swarm mode集群后不允许修改主机名

前提条件

1. 安装Docker Engine 1.12或更新版本
2. 允许2377的tcp端口用于集群管理交互
3. 允许7946的TCP/UDP端口用于节点间的交互（容器网络发现）
4. 允许4789的UDP端口用于overlay网络类型

创建swarm mode集群

[root@swarm-manager ~]#  docker swarm init --advertise-addr 172.16.100.20
Swarm initialized: current node (sc21k9597zasfjaf6cfpuyvy6) is now a manager.

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

    docker swarm join --token SWMTKN-1-10rnutvx6cpja7wv88k7ydywpvjjz1xsj88on00s43te740xca-1pwd9juzgpwnxlxne7p2g93va 172.16.100.20:2377

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

[root@swarm-manager ~]# docker info |grep -A5 Swarm
Swarm: active
 NodeID: pg6fteetxsezu2ygyd3b0joye
 Is Manager: true
 ClusterID: yb5c85p7o054sxp1hb8ieqw43
 Managers: 1
 Nodes: 1
[root@swarm-manager ~]# docker node ls
ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
[root@swarm-manager ~]# netstat -lntp|grep docker
tcp6       0      0 :::2377                 :::*                    LISTEN      1249/dockerd
tcp6       0      0 :::7946                 :::*                    LISTEN      1249/dockerd

添加节点到swarm mode集群

swarm mode集群有manager和worker节点，可通过docker swarm join-token [manager|worker]命令获取节点添加命令

[root@swarm-manager ~]# docker swarm join-token manager
To add a manager to this swarm, run the following command:

    docker swarm join --token SWMTKN-1-5vp5axn28a2cbrtzxlirktbhpnluayacuw81zqacooe3ooe2o3-6ys543fe4zkeagkoaacgaqe3e 172.16.100.20:2377

[root@swarm-manager ~]# docker swarm join-token worker
To add a worker to this swarm, run the following command:

    docker swarm join --token SWMTKN-1-5vp5axn28a2cbrtzxlirktbhpnluayacuw81zqacooe3ooe2o3-64gphy50682jszwc19nn0onpc 172.16.100.20:2377

分别在node1和node2节点上执行如下的docker swarm join命令添加worker节点

[root@swarm-node1 ~]# docker swarm join --token SWMTKN-1-5vp5axn28a2cbrtzxlirktbhpnluayacuw81zqacooe3ooe2o3-64gphy50682jszwc19nn0onpc 172.16.100.20:2377

管理swarm mode集群节点

[root@swarm-manager ~]# docker node ls
ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
lati2179dcwgkvvkc0qcieoim     swarm-node2         Ready               Active
pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
y83k6khc3vxmch1qd3j8kl4ak     swarm-node1         Ready               Active

升/降级节点

• 升级worker节点为manager节点

  [root@swarm-manager ~]# docker node promote swarm-node1 swarm-node2
  Node swarm-node1 promoted to a manager in the swarm.
  Node swarm-node2 promoted to a manager in the swarm.
  # docker node ls
  ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
  lati2179dcwgkvvkc0qcieoim     swarm-node2         Ready               Active              Reachable
  pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
  y83k6khc3vxmch1qd3j8kl4ak     swarm-node1         Ready               Active              Reachable

• 降级manager节点为worker节点

  [root@swarm-manager ~]# docker node demote swarm-node1 swarm-node2
  Manager swarm-node1 demoted in the swarm.
  Manager swarm-node2 demoted in the swarm.
  [root@swarm-manager ~]# docker node ls
  ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
  lati2179dcwgkvvkc0qcieoim     swarm-node2         Ready               Active
  pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
  y83k6khc3vxmch1qd3j8kl4ak     swarm-node1         Ready               Active

  移除节点

  移除节点时需要先在worker节点上执行docker swarm leave命令将节点状态设为Down后，在manager节点上执行docker node rm <node-name>移除。如果要移除manager节点，不建议使用--force强制移除，而应该先进行降级后再移除。

[root@swarm-manager ~]# docker swarm leave
Error response from daemon: You are attempting to leave the swarm on a node that is participating as a manager. Removing the last manager erases all current state of the swarm. Use `--force` to ignore this message.
[root@swarm-manager ~]# docker node rm swarm-node1
Error response from daemon: rpc error: code = 9 desc = node y83k6khc3vxmch1qd3j8kl4ak is not down and can't be removed

[root@swarm-node2 ~]# docker swarm leave
Node left the swarm.
[root@swarm-manager ~]# docker node ls
ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
lati2179dcwgkvvkc0qcieoim     swarm-node2         Down                Active
pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
y83k6khc3vxmch1qd3j8kl4ak     swarm-node1         Ready               Active
[root@swarm-manager ~]# docker node rm swarm-node2
swarm-node2
[root@swarm-manager ~]# docker node ls
ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS
pg6fteetxsezu2ygyd3b0joye *   swarm-manager       Ready               Active              Leader
y83k6khc3vxmch1qd3j8kl4ak     swarm-node1         Ready               Active