RabbitMQ cluster on k8s suitable for OpenStack control plane workloads

Choices and tradeoffs are explained in detail at https://www.mirantis.com/blog/clustered-rabbitmq-kubernetes/

Steps to reproduce the setup on the minikube

• Start minikube with you preferred defults

  minikube start --cpus=2 --memory=4096 --vm-driver=kvm
      

• Create a namespace so you can easily reset an environment when something goes wrong

  kubectl create namespace demo
      

• Make some etcd installation available under etcd name

  kubectl run etcd --image=microbox/etcd --port=4001 --namespace=demo -- --name etcd
  kubectl --namespace=demo expose deployment etcd
      

• Build a Docker image

  eval $(minikube docker-env)
  docker build . -t rabbitmq-autocluster
      

• Initialize k8s secret for erlang cookie

  kubectl create secret generic --namespace=demo erlang.cookie --from-file=./erlang.cookie
      

• Deploy RabbitMQ cluster

  kubectl create -f rabbitmq.yaml
      

• Check that clustering indeed worked

  FIRST_POD=$(kubectl get pods --namespace demo -l 'app=rabbitmq' -o jsonpath='{.items[0].metadata.name }')
  kubectl exec --namespace=demo $FIRST_POD rabbitmqctl cluster_status
      

  And we should see some output similar to the following one (where the most important bit of information is that nodes and running_nodes both contain 3 names):

  Cluster status of node '[email protected]' ...
  [{nodes,[{disc,['[email protected]','[email protected]',
                  '[email protected]']}]},
   {running_nodes,['[email protected]','[email protected]','[email protected]']},
   {cluster_name,<<"rabbit@rabbitmq-deployment-861116474-cmshz">>},
   {partitions,[]},
   {alarms,[{'[email protected]',[]},
            {'[email protected]',[]},
            {'[email protected]',[]}]}]