将manager降级为worker

 

这个问题，说来挺有意思的，我在集群里面创建了2个manager，然后，模拟将第2个manager节点，从集群中移出去，结果发现报错了：

 

[root@nccztsjb-node-07 ~]# docker node rm cw5bome4a9g1rzelivxejd31c

Error response from daemon: rpc error: code = FailedPrecondition desc = node cw5bome4a9g1rzelivxejd31c is a cluster manager and is a member of the raft cluster. It must be demoted to worker before removal

 

 

报错信息说：在移除该节点是manager节点，在移除之前，必须首先降级为worker节点。

 

降级的方法也很简单：

 

docker node demote cw5bome4a9g1rzelivxejd31c

 

 

使用node的demote参数，后面加上node的ID 【通过docker node ls可以查看到】

[root@nccztsjb-node-07 ~]# docker node demote cw5bome4a9g1rzelivxejd31c
Manager cw5bome4a9g1rzelivxejd31c demoted in the swarm.

 

 

降级后，可以看到，原来在MANAGER STATUS列中的Reachable状态已经没有了，说明是个worker节点。

[root@nccztsjb-node-07 ~]# docker node ls
ID                            HOSTNAME           STATUS    AVAILABILITY   MANAGER STATUS   ENGINE VERSION
mjktd9is5xd8d4el2awrqby5k *   nccztsjb-node-07   Ready     Active         Leader           20.10.17
u87i9artgcnyki0osp9gvylu3     nccztsjb-node-08   Ready     Active                          20.10.17
z1x3ufsdxekrt6w5n37m7azfe     nccztsjb-node-09   Ready     Active                          20.10.17
cw5bome4a9g1rzelivxejd31c     nccztsjb-node-10   Ready     Active                          20.10.17
 

 

manger降级之后，就可以对其进行remove的操作了。当然remove之前要先drain才行。

 

将worker升级为manager

 

那反过来，将worker升级为manager 呢？

 

docker node promote cw5bome4a9g1rzelivxejd31c

 

 

[root@nccztsjb-node-07 ~]# docker node promote cw5bome4a9g1rzelivxejd31c
Node cw5bome4a9g1rzelivxejd31c promoted to a manager in the swarm.
[root@nccztsjb-node-07 ~]# docker node ls
ID                            HOSTNAME           STATUS    AVAILABILITY   MANAGER STATUS   ENGINE VERSION
mjktd9is5xd8d4el2awrqby5k *   nccztsjb-node-07   Ready     Active         Leader           20.10.17
u87i9artgcnyki0osp9gvylu3     nccztsjb-node-08   Ready     Active                          20.10.17
z1x3ufsdxekrt6w5n37m7azfe     nccztsjb-node-09   Ready     Active                          20.10.17
cw5bome4a9g1rzelivxejd31c     nccztsjb-node-10   Ready     Active         Reachable        20.10.17
[root@nccztsjb-node-07 ~]# 

 

 

这样，就将worker节点升级为manager节点了。

 

标签：node,worker,swarm,manager,nccztsjb,cw5bome4a9g1rzelivxejd31c,Ready,docker,节点
来源： https://www.cnblogs.com/chuanzhang053/p/16649626.html