Other options -
https://github.com/nanit/kubernetes-rabbitmq-cluster
Experiment to produce a "kubernentes first" rabbitmq cluster manager.
- zero user effort high availablity and fault tolerence.
Http request to
http://localhost:15672/api/vhosts to collect list of vhosts
then for each vhost -
http://localhost:15672/api/aliveness-test/%2F
from http://hg.rabbitmq.com/rabbitmq-management/raw-file/rabbitmq_v2_2_0/priv/www-api/help.html
Failure means sidekick dies, then pod dies and is replicated somewhere else
This will need an admin user's credentials which will need to bootstrapped in the rabbitmq docker image.
- rabbitmq-cluster-sidekick - handles dynamic cluster managment.
- rabbitmq-cluster-admin - a clusterwide ui for managment of the cluster(s).
- docker images
- kubernetes yaml configuration
rabbitmq-cluster-node - hosts a rabbitmq instance and rabbitmq-cluster-sidekick instance
rabbitmq-cluster-admin - hosts a rabbitmq-cluster-admin instance
rabbitmq-cluster-rc.yaml
rabbitmq-cluster-service.yaml
rabbitmq-cluster-admin-rc.yaml
rabbitmq-cluster-admin-service.yaml
- kubernentes
- docker
- etcd
- golang