RabbitMQ 3.6 Rancher Docker image

This template creates and scales out a rabbitmq-3.6 cluster.

How it works

The entrypoint calls confd to create /etc/rabbitmq/rabbitmq.config file; in the cluster_nodes directive lists all the running rabbitmq containers so that the node connects to the others and creates or joins the cluster at startup time.

After deploying the first container, you can scale the service adding new container via Rancher UI.

To access the management interface, point a balancer on the 15672 port of this service.

Environment variables

The following environment variables are passed to confd in order to set up RabbitMQ's configuration file:

• RABBITMQ_CLUSTER_PARTITION_HANDLING: RabbitMQ's cluster handling setting: default set to autoheal
• RABBITMQ_NET_TICKTIME: adjusts the frequency of both tick messages and detection of failures: default set to 60
• RABBITMQ_ERLANG_COOKIE: cookie to allow nodes communication: default set to defaultcookiepleasechange

Other two variables are available to fine-tune the cluster or test confd configuration:

• ALTERNATE_CONF: overrides the whole default confd RabbitMQ template: default set to empty
• CONFD_ARGS: additional confd args along with default --backend rancher --prefix /2015-07-25: default set to --interval 5