Setup REDIS service with VRRP

Redis service

Provides a common cache, required if clustered mode is in use for NtkAS or any service is balanced using an external load balancer.

Start the REDIS server on every node:

ntk redis start

To check that the service is running, inspect the cache:

docker exec -it redis redis-cli -n 1
127.0.0.1:6379[1]> keys *
  1) "main-dashboard/cache:a8691165dececc2214a89b52fa9672068558bfee"
  2) "main-dashboard/cache:9c61bbcf-16f5-4b8e-9329-a6c15f62bf53"
 ....

Use the REDIS keys * command to ensure that the cache is being populated. This has to be done after NtkAS cache configuration and ntk as restart.

VRRP VIP and tracker configuration

Configure the tracker on every node in the VRRP cluster:

ntk cfg set vrrp.track.tcp '{
      "host": "127.0.0.1",
      "port": "6379",
      "timeout": "2"
    }' --json-input

Configure the redis VRRP service on each node with VRRP running. Ensure address and vrid are the same and prio is different on each node. Replace vrid and address with your local config.

ntk cfg set vrrp.services.redis '{
      "prio": "150",
      "vrid": "53",
      "interface": "eth0",
      "address": "10.0.1.98",
      "preempt": "off",
      "track": [
        "tcp"
      ]
    }' --json-input

Restart the VRRP service on all VRRP nodes:

ntk vrrp restart