RabbitMQ monitoring Consumption rate/ Re-delivery rate via Prometheus and Grafana showing different numbers

[David Ohayon]

Hey friends. 

We're currently setting up our monitoring stack for our RMQ setup, however, we can't seem to get the numbers to correlate. 

For example, when checking for redelivery rate or delivery rate, the numbers are completely different from what we find in the management UI.

For example, the following Prom query:
sum by (queue) (rate(rabbitmq_channel_messages_redelivered_total{job="rabbitmq-exporter", queue=~"$rmq_queue_prefix.*"}[5m])) 

We are getting a redelivery rate for a certain queue of 0.8/s, where the management UI reports 5-8 messages per second. Is this sort of discrepancy expected? How to deal with this?

Thanks!