验证Redision是可以自动拓扑刷新的

1.首先检查当前集群节点的状态

192.168.1.56(192.168.1.56:6579)>cluster nodes "db73aebc13f801f3e7682f3020dda9a2de78a2de 192.168.1.56:6679@16679 master - 0 1657246502566 94 connected 10923-16383 42a9149f798f415ef1b240075600d9897805ada1 192.168.1.56:6680@16680 slave 58318a93a1f87019af06fa96ab9692e182c9184d 0 1657246502000 88 connected 6b8eec08fdf24c9a72a55fcbe4a9405458b703e8 192.168.1.56:6480@16480 slave db73aebc13f801f3e7682f3020dda9a2de78a2de 0 1657246502364 94 connected 58318a93a1f87019af06fa96ab9692e182c9184d 192.168.1.56:6579@16579 myself,master - 0 1657246501000 88 connected 5461-10922 e18554b6ad8a8d78e41e8e2c225d8aaba6c8ea5c 192.168.1.56:6580@16580 slave bd77823c2ebe3fdeea7ce9f58fa0a814024492ec 0 1657246502566 93 connected bd77823c2ebe3fdeea7ce9f58fa0a814024492ec 192.168.1.56:6479@16479 master - 0 1657246501358 93 connected 0-5460 

翻译是： 192.168.1.56:6679 主节点 192.168.1.56:6680 从节点来看，对应的主要是192.168.1.56:6579 192.168.1.56:6480 从节点，对应的主是192.168.1.56:6679 192.168.1.56:6579 主节点 192.168.1.56:6580 从节点来看，对应的主要是192.168.1.56:6479 192.168.1.56:6479 主节点

2、创建Redis压测接口

@GetMapping("/test/redis")     public String test() throws InterruptedException { 
                ////模拟业务处理耗时         Thread.sleep(10);         if(!RedisUtils.hasKey("test")){ 
                    RedisUtils.setStrWithTime("test","test",100);         }         return RedisUtils.getStr("test");     } 

3、现在使用Jmeter压测Redis集群，正常。 4、现在手动down掉 192.18.1.56:6479 节点

docker stop redis_m6479

5、查看当前集群状态

"db73aebc13f801f3e7682f3020dda9a2de78a2de 192.168.1.56:6679@16679 master - 0 1657251138156 94 connected 10923-16383
42a9149f798f415ef1b240075600d9897805ada1 192.168.1.56:6680@16680 slave 58318a93a1f87019af06fa96ab9692e182c9184d 0 1657251140169 88 connected
6b8eec08fdf24c9a72a55fcbe4a9405458b703e8 192.168.1.56:6480@16480 slave db73aebc13f801f3e7682f3020dda9a2de78a2de 0 1657251140571 94 connected
58318a93a1f87019af06fa96ab9692e182c9184d 192.168.1.56:6579@16579 myself,master - 0 1657251140000 88 connected 5461-10922
e18554b6ad8a8d78e41e8e2c225d8aaba6c8ea5c 192.168.1.56:6580@16580 master - 0 1657251140000 95 connected 0-5460
bd77823c2ebe3fdeea7ce9f58fa0a814024492ec 192.168.1.56:6479@16479 master,fail - 1657251082816 1657251080501

发现192.168.1.56:6479 显示fail，192.168.1.56:6580自动故障切换为主节点。 同时 Redission自动感知拓扑刷新，控制台打印如下日志：

2022-07-08 11:33:15,748 | INFO | o.r.c.MasterSlaveEntry - master 192.168.1.56/192.168.1.56:6479 used as slave 
2022-07-08 11:33:15,748 | INFO | o.r.c.ClusterConnectionManager - slave redis://192.168.1.56:6580 removed for slot ranges: [[0-5460]] 
2022-07-08 11:33:16,078 | INFO | o.r.c.p.MasterPubSubConnectionPool - 1 connections initialized for 192.168.1.56/192.168.1.56:6580 
2022-07-08 11:33:16,862 | INFO | o.r.c.ClusterConnectionManager - 192.168.1.56/192.168.1.56:6479 master and related slaves: [addr=redis://192.168.1.56:6580] removed 
2022-07-08 11:33:16,862 | INFO | o.r.c.p.MasterConnectionPool - 24 connections initialized for 192.168.1.56/192.168.1.56:6580 
2022-07-08 11:33:17,051 | INFO | o.r.c.p.PubSubConnectionPool - 1 connections initialized for 192.168.1.56/192.168.1.56:6580 
2022-07-08 11:33:17,741 | INFO | o.r.c.ClusterConnectionManager - master: redis://192.168.1.56:6580 added for slot ranges: [[0-5460]] 
2022-07-08 11:33:17,741 | INFO | o.r.c.p.SlaveConnectionPool - 24 connections initialized for 192.168.1.56/192.168.1.56:6580 

再次压测Redis集群，集群正常。