Redis sentinel deployment explained

SHORT VERSION:

First we build the images:

docker-compose -f docker-compose-redis.yml build --no-cache redis-sentinel-alp

Then we start them:

docker-compose -f docker-compose-redis.yml scale redis-sentinel-alp=3

Cahnge the 3 number for the number of nodes needed

---

LONG VERSION

THE IMAGES:

• redis-sentinel-alp
  • Standard redis 3.2.12 image;
  • Dockerfile: copies redis.conf and a start script;
  • docker-compose-redis.yml: docker-compose file;
  • redis.conf: redis configuration;
  • sentinel.conf: sentinel2 configuration;
  • start-redis.sh: script to configurate the nodes;

HOW TO START EVERYTHING:

Then we start them. NUMBER is the number of redis instances you want:

docker-compose -f docker-compose-redis.yml scale redis-sentinel-alp=<NUMBER>

Configurations:

• In docker-compose file, redis-cluster-manager image, environment variables:
  • prefix_network: prefix of the docker network to config nodes;
  • init_ip: initial last number ip of nodes to use in the process to select master node;
  • last_ip: last number ip of nodes to use in the process to select master node;

When containers are started, the script find ips availables in the range specified using prefix_network, init_ip and last_ip params, and set the node with the lowest ip with the master role, and other are slaves nodes of the master.

Checking if everything is alright

The redis containers will be showing everything that happens on them in their console windows.

You can also connect to one of the containers

docker exec -it container_id /bin/bash