Jedis connection in block state

[sagarsitap596]

I have JedisCluster with 3 shards and 1 repilica each with below pool

final JedisPoolConfig poolConfig = new JedisPoolConfig();
poolConfig.setMaxTotal(200);
poolConfig.setMaxIdle(200);
poolConfig.setMinIdle(16);
return poolConfig;

My application is making around 5-7 lacs of request per minute to redis (insert/update/delete key) over multiple threads (15-20).

After some time one of jedis connection thread doesn't respond. When I trace down the thread trace I found :

java.util.concurrent.locks.abstractqueuedsynchronizer$conditionobject.await()

Is it because jedis connection pool is exausted?
Do I need to increase connection pool size ?

[sazzad16]

It seems to be related to:

• either JedisPool exhausted in Jedis 2.10.0 #1920 which can be resolved by upgrading Jedis version
• or JedisCluster hangs entire threads sharing the same JedisCluster when the redis is temporarily overloaded. #1620 which can be resolved by a positive number in setMaxWaitMillis of pool config