jedis-3.2.0 JedisClusterMaxAttemptsException connect to redis-5.0.7 cluster

[DoverDee]

Expected behavior

when kill 7000 master node , the code will appear exception,
wait a moment, the slave instead of 7000 master node, the exception should be disappear.
but still pose exception .

Java code

public static void clusterConnJedis() {
Set clusterNodes = new HashSet();
HostAndPort hp7000 = new HostAndPort("172.16.1.41", 7000);
HostAndPort hp7001 = new HostAndPort("172.16.1.41", 7001);
HostAndPort hp7002 = new HostAndPort("172.16.1.41", 7002);
HostAndPort hp7003 = new HostAndPort("172.16.1.41", 7003);
HostAndPort hp7004 = new HostAndPort("172.16.1.41", 7004);
HostAndPort hp7005 = new HostAndPort("172.16.1.41", 7005);
clusterNodes.add(hp7000);
clusterNodes.add(hp7001);
clusterNodes.add(hp7002);
clusterNodes.add(hp7003);
clusterNodes.add(hp7004);
clusterNodes.add(hp7005);
JedisCluster clusterPool = null;
while (true) {
try {
clusterPool = new JedisCluster(clusterNodes, 5000, new JedisPoolConfig());
int index = new Random().nextInt(100000);
String key = "k-" + index;
String value = "v-" + index;
clusterPool.set(key, value);
System.out.println("key:value=" + key + ":" + clusterPool.get(key));
TimeUnit.MILLISECONDS.sleep(500);
} catch (Exception e) {
e.printStackTrace();
} finally {
if (null != clusterPool) {
clusterPool.close();
}
}
}
}

Actual behavior and Steps to reproduce:

Redis cluster just installed and runing, run TestClass.clusterConnJedis(); in main method. everything is ok , I find the right pint key:value=k-RandomNum:v-RandomNum per 500 MILLISECONDS in conole.

But when I command ' kill -p ${hp7000_ pid}' , 7000 master node be killed. I find the console print
JedisClusterMaxAttemptsException several normal print intervals.
exception details:
redis.clients.jedis.exceptions.JedisClusterMaxAttemptsException: No more cluster attempts left.
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:86)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:124)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:124)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:124)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:124)
at redis.clients.jedis.JedisClusterCommand.runWithRetries(JedisClusterCommand.java:124)
at redis.clients.jedis.JedisClusterCommand.run(JedisClusterCommand.java:25)
at redis.clients.jedis.JedisCluster.set(JedisCluster.java:143)
at Test.clusterConnJedis(Test.java:69)
at Test.main(Test.java:11)
exception picture:

Redis / Jedis Configuration

see code clusterConnJedis()

Jedis version:

jedis-3.2.0

Redis version:

redis-5.0.7 cluster

Java version:

jdk1.8.0.211

[yangbodong22011]

When the Cluster is abnormal, Jedis will provide retries. The default number of retries is 5, see redis.clients.jedis.BinaryJedisCluster#DEFAULT_MAX_ATTEMPTS. You can specify more retries through the following construction method.

public JedisCluster(Set<HostAndPort> nodes, int timeout, int maxAttempts) {
    this(nodes, timeout, maxAttempts, new GenericObjectPoolConfig());
}

// your code:
clusterPool = new JedisCluster(clusterNodes, 5000, maxAttempts);

But I don't recommend that you set maxAttempts too large, otherwise it may cause java.lang.StackOverflowError due to recursion.

[sazzad16]

@DoverDee What is the output of CLUSTER NODES from redis-cli. I'm guessing node IPs are not with 172.16.1.41.

[sazzad16]

Closed due to inactivity.

[walles]

This PR would provide retries with backoff, giving the cluster a chance to come back: #2358

[choice]

May I ask how this problem was finally dealt with.
'redis.clients.jedis.exceptions.JedisClusterMaxAttemptsException: No more cluster attempts left.'
This problem is giving me a headache

[jbhamilton]

Running into the same issue in our production env right now and cannot replicate locally. Any guidance would be appreciated.