scale rabbitmq stream consumption using super streams partitioned across cluster nodes

Steps

1. Create a super stream with the rabbitmq-streams add_super_stream command and a partition count, which creates that many individual partition streams under the hood.
2. Publish using a super-stream-aware client, which routes each message to the correct partition stream based on a routing key or hash.
3. Consume using the super stream client APIs so consumers get assigned partitions automatically, similar to Kafka-style partition assignment.
4. Rely on single active consumer semantics per partition to preserve ordering within each partition while still parallelizing across partitions.
5. Reserve super streams for cases where a single stream's throughput or storage genuinely becomes a bottleneck rather than defaulting to them for all streams.