Setting up a Redis cluster without replication is possible but not normally recommended due to potential loss of data resilience. However, if you decide to go this route, it essentially means that each node in the cluster will hold independent data and no nodes will serve as slaves.
Here's a high level process of how you can do it:
redis-cli tool to create the cluster after all the instances are up and running. The following example creates a cluster with nodes running on ports 7000, 7001 and 7002:
During the creation process, the tool will ask how to distribute hash slots among nodes. By default, it offers an equal distribution. Accept this for most use cases.
Then in the CLI, use the 'CLUSTER NODES' command:
If your cluster is correctly set up, it will display information about each node in the cluster.
Please note that running a Redis cluster without replication means your data won't have the redundancy typically provided by slave nodes. This makes your cluster more vulnerable to data loss if one or more nodes fail.