I've implemented the example in examples/reconnect_to_cluster_nodes to try to reconnect at the beginning of the program, but in a situation where rabbitmq terminates during a normal consume, the program terminates with a memory-related exception.

Is it possible to implement a feature to recover from such a situation?