操作步骤
查看当前的集群情况
1 2 3 4 5 6 7 8 9 root@knight:/data# redis-cli -p 6381 -c 127.0.0.1:6381> cluster nodes 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 192.168.1.42:6383@16383 master - 0 1691996146001 3 connected 10923-16383 849a82f0bd85238762ca7ccf234aa30ba97d93a0 192.168.1.42:6386@16386 slave f999e35136ec2e61fcceebf182f5c38ef4a4354d 0 1691996144999 1 connected 9f55acb1e18b1dfa53ad3a4a506ff2512a92cb46 192.168.1.42:6385@16385 slave 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 0 1691996144000 3 connected f999e35136ec2e61fcceebf182f5c38ef4a4354d 192.168.1.42:6381@16381 myself,master - 0 1691996145000 1 connected 0-5460 8e70d538b699c96edc3f6608224aa8ecf55d857b 192.168.1.42:6384@16384 slave 6dca84998e245ce4cc6c92882fb7ac94d501efda 0 1691996144000 2 connected 6dca84998e245ce4cc6c92882fb7ac94d501efda 192.168.1.42:6382@16382 master - 0 1691996144000 2 connected 5461-10922 127.0.0.1:6381>
当前登录主机为1号机，6号机是1号机的从服务器，5号机是3号机的从服务器，4号机是2号机的从服务器。
关闭1号机容器，然后再次查看容器的状态
关闭1号容器
1 docker stop redis-node-1
效果：
折叠代码 1 2 3 4 5 6 7 8 9 10 root@knight:/docker# docker stop redis-node-1 redis-node-1 root@knight:/docker# docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 61d59245db3d redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Up 5 hours redis-node-6 223fad42e069 redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Up 5 hours redis-node-5 f0055941fe3f redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Up 5 hours redis-node-4 e361d9695a0c redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Up 5 hours redis-node-3 400119c3015c redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Up 5 hours redis-node-2 61e2e38927c4 redis:6.0.8 "docker-entrypoint.s…" 28 hours ago Exited (0) 9 seconds ago redis-node-1
可以看到1号容器已经被关掉了。
然后以2号机作为切入点，查看集群的状态。
进入2号容器：
1 docker exec -it redis-node-2 bash
效果：
1 2 3 4 5 6 7 8 9 10 root@knight:/docker# docker exec -it redis-node-2 bash root@knight:/data# redis-cli -p 6382 -c 127.0.0.1:6382> cluster nodes 8e70d538b699c96edc3f6608224aa8ecf55d857b 192.168.1.42:6384@16384 slave 6dca84998e245ce4cc6c92882fb7ac94d501efda 0 1692013667508 2 connected 6dca84998e245ce4cc6c92882fb7ac94d501efda 192.168.1.42:6382@16382 myself,master - 0 1692013666000 2 connected 5461-10922 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 192.168.1.42:6383@16383 master - 0 1692013667000 3 connected 10923-16383 f999e35136ec2e61fcceebf182f5c38ef4a4354d 192.168.1.42:6381@16381 master,fail - 1692013443386 1692013440000 1 disconnected 9f55acb1e18b1dfa53ad3a4a506ff2512a92cb46 192.168.1.42:6385@16385 slave 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 0 1692013665000 3 connected 849a82f0bd85238762ca7ccf234aa30ba97d93a0 192.168.1.42:6386@16386 master - 0 1692013665498 7 connected 0-5460 127.0.0.1:6382>
可以看到当前6号机处于master状态，顶替了原来1号机的位置，而1号机现在处于fail状态。
这就完成了redis主从切换的演示，补充一句，如果把1号机重新启动回来的话，也不会影响现在6号机的master状态，1号机会作为6号机的slave机器。
效果如下：
折叠代码 1 2 3 4 5 6 7 8 9 10 11 12 root@knight:/docker# docker start redis-node-1 redis-node-1 root@knight:/docker# docker exec -it redis-node-1 bash root@knight:/data# redis-cli -p 6381 -c 127.0.0.1:6381> cluster nodes 849a82f0bd85238762ca7ccf234aa30ba97d93a0 192.168.1.42:6386@16386 master - 0 1692013947000 7 connected 0-5460 f999e35136ec2e61fcceebf182f5c38ef4a4354d 192.168.1.42:6381@16381 myself,slave 849a82f0bd85238762ca7ccf234aa30ba97d93a0 0 1692013947000 7 connected 9f55acb1e18b1dfa53ad3a4a506ff2512a92cb46 192.168.1.42:6385@16385 slave 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 0 1692013946019 3 connected 8e70d538b699c96edc3f6608224aa8ecf55d857b 192.168.1.42:6384@16384 slave 6dca84998e245ce4cc6c92882fb7ac94d501efda 0 1692013949036 2 connected 6dca84998e245ce4cc6c92882fb7ac94d501efda 192.168.1.42:6382@16382 master - 0 1692013947022 2 connected 5461-10922 41f99f5142b5c4b7e1e17c8a94e88fc74cda4d40 192.168.1.42:6383@16383 master - 0 1692013948030 3 connected 10923-16383 127.0.0.1:6381>