It's normal that you observe latency when retrieving data in a jedis cluster, depending on the clustering strategy you have, because you will receive money from one jedis to another to take data. The red arrow is the reason for adding time to queries in the jedis cluster.

The only way to access the correct instance is to understand by key the location of your data, enter into the key, for example, the location: "instance3 / user / 5" or find an instance of the location that performs calculations on the keys.
source
share