docker swarm mode 之二 --- 集群的自动迁移和故障恢复

docker swarm mode 集群的自动迁移和故障恢复

测试自动迁移

在swarm3停掉docker

[root@swarm3 ~]# systemctl stop docker

查看集群上stack

[root@swarm2 overlay]# docker stack ls
NAME        SERVICES
testlink    1
jenkins     1
harbor      6
mantis      1
prometheus  6

显示当前所有stack的containers，发现已经没有运行在swarm3上面的容器了

[root@swarm3 overlay]# for i in `docker stack ls |tail -5 |awk '{print $1}'`;do docker stack ps $i |awk '{if(($5=="DESIRED")||($5=="Running")) printf "%12s%10s%10s\n", $1,$4,$5}'; done
      ID      NODE   DESIRED
sfo2qke6zltx    swarm1   Running
      ID      NODE   DESIRED
8bz53euw4cru    swarm2   Running
kced38vhycgm    swarm2   Running
o74beuybld7j    swarm2   Running
q0kfoykzuci7    swarm1   Running
l30scvwcj02n    swarm1   Running
t1swltc5o3rp    swarm1   Running
2xnzkh2ft4jc    swarm2   Running
imq6r6gqxa7q    swarm1   Running
pau6cwtev948    swarm1   Running
      ID      NODE   DESIRED
ojw1uv2li5zu    swarm2   Running
      ID      NODE   DESIRED
74hjcvm1ezgq    swarm1   Running
      ID      NODE   DESIRED
ytprsour0xlk    swarm2   Running
m2ffry5j6etw    swarm2   Running
gmzlueiq3j0v    swarm2   Running
l86xnc4ukoa7    swarm1   Running
i4jpnehg73g1    swarm1   Running
uq7jqj90awl0    swarm1   Running