Understand the default cluster configuration
The Bitnami Multi-Tier Solution for RabbitMQ uses multiple VMs, consisting of a broker of one or more nodes, to provide a horizontally scalable RabbitMQ cluster where all data/state is replicated across all nodes. The cluster is set with the configuration below:
- Network partition is se to ignore mode as it is recommended for reliable networks. Refer to RabbitMQ Partitions Documentation for more information about partitions.
- The are no mirrored queues.
- Shovel and Federation plugins are not used.
- All nodes are disk nodes. Refer to RabbitMQ Clustering Documentation for more information about the kind of nodes.
- Disk free limit is set to mem_relative, 1.0. Refer to RabbitMQ Disk Alarms Documentation about Disk Alarms.
There are other configuration for the RabbitMQ cluster such as using Mirrored Queues. Please refer to RabbitMQ HA Documentation for more information about these alternative solutions.