Fix for issue #39. Cluster setting cluster_partition_handling can now…

[wasfree]

This is the pull request for #39 and will also fix #37. Cluster setting cluster_partition_handling can now be changed via pillars. In regular the default value for cluster_partition_handling is ignore (see https://www.rabbitmq.com/configure.html#configuration-file), however this formula used autoheal as default in the past. So i would recommend to keep it on autoheal (see change in map.jinja), if a change of this value is need this can now be done via pillars.

example pillars:

rabbitmq:
  server:
    enabled: true
    bind:
      address: 0.0.0.0
      port: 5672
    secret_key: rabbit_master_cookie
    admin:
      name: adminuser
      password: pwd
    memory:
      vm_high_watermark: 0.4
    plugins:
      - amqp_client
      - rabbitmq_management
  cluster:
    enabled: true
    role: master
    mode: disc
    partition_handling: ignore
    members:
    - name: openstack1
      host: 10.10.10.212
    - name: openstack2
      host: 10.10.10.213

[wasfree]

Hi, sorry had to close #40 and reopened it in #41. Pillar was changed from cluster_partition_handling to partition_handling. Can someone please look into this merge request? Thanks!

[fpytloun]

Great, thank you for contribution :-)