Connecting to redis cluster using Java/Python clients

This story is in continuation of my 3 parts redis series. First — Creating a redis cluster. Second — Horizontally scaling in/out a redis cluster and lastly this story which will explain how can we connect to a redis cluster from within our java or python applications.

Using Java

Let us first see how can we connect to a redis cluster using Java. For this purpose i created a spring boot project. We will be using jedis client to connect to redis cluster. Know more about Jedis here. The dependencies tag of my pom.xml looks something like as follows-

pom.xml dependencies

application.properties file under resources directory should look like as follows-

application.properties

Notice that i haven’t set a password for simplicity. You can use a password if your redis cluster is password protected.

Next we would want to create a Bean using class JedisCluster of package redis.clients.jedis to establish a connection with redis cluster which will use values in application.properties file as shown in following gist-

Configuration class to establish connection with redis cluster

In the above gist, we provided only 1 host i.e localhost with port 6001 to connect to, However when connection is established JedisCluster gathers information about other nodes in the cluster. So even if localhost:6001 goes down, it will still be able to communicate with the cluster through some other node eg. 6002 or 6003 etc. This essentially means that JedisCluster manages failover scenarios. Until and unless there is an operational cluster, JedisCluster will be able to connect to it.

We can now Autowire JedisCluster class in our application’s DAO layer to interact with redis cluster as shown below-

@Component
public class RedisDaoImpl {
   
   @Autowired
   private JedisCluster getRedisCluster;

   public void setKeyInRedis(String key, String value) {
      getRedisCluster.set(key, value);
   }

   public String getValueByKey(String key) {
      return (String) getRedisCluster.get(key);
   }
}

setKeyInRedis() will set a key in redis and getValueByKey() will get the value of key. Private autowired variable getRedisCLuster will now be capable to do a host of operations apart from just get() and set() eg. — exists() , persist() , expire() and many more redis operations.

Using Python

To connect a python application to redis cluster we will use redis-py-cluster. Know more about redis-py-cluster here.

Firstly you need to install redis-py-cluster

$ pip install redis-py-cluster

All we need to do now is import RedisCLuster and define some configuration.

See gist below-

Variable rc will now be capable to do all redis operations like exists(), expire() etc.

It is recommended to add all nodes in the cluster in List startup_nodes. Even if you don’t, it will still identify all nodes in cluster on its own and will manage failover scenarios. This means that in above scenario, if 127.0.0.1:6001 goes down, it will still be able to connect to redis cluster without us explicitly telling it which all nodes are part of the cluster.

This marks the end of 3 part redis series where we learnt how to create a redis cluster then we learnt how to horizontally scale in/out a redis cluster and after this story we also know how to connect to a redis cluster programatically.

Enjoy redis…!!!