Machine Learning & Big Data Blog

Redis Clustering & Partitioning for Beginners

3 minute read
Walker Rowe

This tutorial introduces you to clustering and partitioning in Redis.

(This tutorial is part of our Redis Guide. Use the right-hand menu to navigate.)

Clustering & partitioning in Redis

One way to boost the performance of Redis is to put all records with the same keys into the same node. In that case only one node needs to be read when looking for values with that key. Here we explain the principles behind that.

Suppose you want to separate customers, employees, and vendors into separate nodes. The way to do that would be to:

  • set up a Redis cluster.
  • assign cluster slots to specific nodes in the cluster.

Then write data for each customer like this, where the hset function sets the key using the hash slot function function:

hset {customer}.dodgedealer car "dodge" year 2016
hset {customer}.forddealer car "ford" year 2016

In this example, both car dealers Dodge and Ford will be assigned to the same hash slot, and thus the same node, since the curly braces {} determines what part of the key is used to calculate the hash slot.

Redis assigns keys to nodes by calculating the hash slot like this:

CRC16(key) mod 16384

CRC16 is a hashing function (i.e., converts a string to a number) and mod 16384 means modulus 16384. Modulus is the integer remainder when dividing an integer by the modulus. For example, 4 mod 2 = 0 because 2 divides 4 evenly. 3 mod 2 = 1 since 3 = (2 * 1 ) + 1.

CRC16(key) mod n will always produce the same result. So it yields a predictable value suitable for this assignment. To see that open a Redis shell and type what is shown below.

The first command adds the hash slot to the node to which you have attached:


Then we show on the screen what the hash slot value of the key walker would be:

(integer) 10528

Now assign a key and write it to this node:

hset walker "walker"
(error) CLUSTERDOWN The cluster is down

We get an error CLUSTERDOWN because I have not yet set up the cluster, only one node in the cluster.

Below we show how to set up nodes in a cluster. If you want to automate the whole process install Ruby and then run this Ruby program following these instructions.

Create Redis Instance

Install Redis following these instructions and then:

Make 3 directories for 3 instances:

mkdir 7001 7002 7003

Into each of these folders copy this redis.conf, changing the port number as you do.

port 7001
cluster-enabled yes
cluster-config-file 7001.nodes.conf
cluster-node-timeout 5000
appendonly yes

Now start each instance:

nohup src/redis-server 7001/redis.conf >7001.log 2>&1&
nohup src/redis-server 7002/redis.conf >7002.log 2>&1&
nohup src/redis-server 7003/redis.conf >7003.log 2>&1&

Check the logs to see each started:

tail 7001.log
19908:M 18 Jan 01:10:27.927 * Ready to accept connections

Now pick one node and connect to it. Notice that since we have not run the Ruby program the nodes are not aware of each other. So this is not a complete cluster install. But it is enough to show you how to set up key partitioning. The Ruby program updates these node files to put in master/slave/replica/nodes and other info, which makes the cluster complete.

redis-cli -h localhost -p 7001



Ask the node what other nodes there are and it shows the configuration only for itself.

cluster nodes
>eeba92cc34ef032da1f5dbcf7450b93aca9bf62f :7001@17001 myself,master - 0 0 0 connected

See below we have not assigned any hash_slot to this node yet:

cluster info

Alternate techniques

There are other ways to do this including using the Redis proxy written by Twitter.

Twitter says theses companies are using it:

  • Twitter
  • Wikimedia
  • Pinterest
  • Snapchat
  • Flickr
  • Yahoo!
  • Tumblr

Learn ML with our free downloadable guide

This e-book teaches machine learning in the simplest way possible. This book is for managers, programmers, directors – and anyone else who wants to learn machine learning. We start with very basic stats and algebra and build upon that.

These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.

See an error or have a suggestion? Please let us know by emailing

Business, Faster than Humanly Possible

BMC works with 86% of the Forbes Global 50 and customers and partners around the world to create their future. With our history of innovation, industry-leading automation, operations, and service management solutions, combined with unmatched flexibility, we help organizations free up time and space to become an Autonomous Digital Enterprise that conquers the opportunities ahead.
Learn more about BMC ›

About the author

Walker Rowe

Walker Rowe is an American freelancer tech writer and programmer living in Cyprus. He writes tutorials on analytics and big data and specializes in documenting SDKs and APIs. He is the founder of the Hypatia Academy Cyprus, an online school to teach secondary school children programming. You can find Walker here and here.