JedisCluster throws Too many Cluster redirections when a master node crash

[wenxuanguan]

I have redis cluster with 3 masters and 3 slaves and when ever I run DEBUG SEGFAULT to crash master node, it throws an exception
code snap shot:
public static void main(String[] args) {
TestCluster tr = new TestCluster();
try {
tr.parseCommandLine(args);
}
}

public class TestCluster {
public void parseCommandLine() {
jedisClusterNodes.add(new HostAndPort("10.0.1.21", 6379));
jedisClusterNodes.add(new HostAndPort("10.0.1.141", 6379));
......//add 6 cluster nodes ip and port
testTPS(nums, concurrent);
}

public void testTPS() {
Thread[] threads = new MyThread[concurrent];
JedisCluster[] jc = new JedisCluster[concurrent];
for(int i=0; i<concurrent; i++){
jc[i] = new JedisCluster(jedisClusterNodes);
threads[i] = new MyThread(jc[i], nums, concurrent);
}
for(int i=0; i<concurrent; i++){
threads[i].start();
}
}
public void doSet() {
for(long i=0; i<nums; i++){
String ret = jc.set(key, value);
}
}
class MyThread extends Thread{
public void run(){
doSet();
}
}
}
.......................................................................
Exception throws when the master crashed
Exception in thread "Thread-1" redis.clients.jedis.exceptions.JedisClusterMaxRedirectionsException: Too many Cluster redirections?
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:34)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:68)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:85)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:68)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:85)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:68)
at redis.clients.jedis.JedisClusterCommand.run(JedisClusterCommand.java:29)
at redis.clients.jedis.JedisCluster.set(JedisCluster.java:96)
at TestCluster.doSet(TestCluster.java:84)
at TestCluster$MyThread.run(TestCluster.java:114)
.............................................................................
after master node crashed, the slave node failover it.
the same thing happened with JedisCluster.get()
test env:
Jedis 2.7.3
redis 3.0.3

Thanks

[zouyx]

@wenxuanguan
hello, i think this issue is the same to yours ,may help you #1238

[trojanliu]

Caused by: redis.clients.jedis.exceptions.JedisClusterMaxRedirectionsException: Too many Cluster redirections

[github-actions]

This issue is marked stale. It will be closed in 30 days if it is not updated.