Hi,
we are currently running RabbitMQ 3 nodes cluster (versions RMQ: 3.8.30, Erland: 23.3.4.11) and we've spotted a very strange behaviour during e2e tests. Messages put to the topic exchange stopped being delivered to selected queues. Naturally we looked at the configuration in the UI, all queues were declared the same way with the same bounding key.
Then we stopped all the consuming microservices and put the message manually to the exchange, same behaviour.
We searched the rabbitmq logs and found no errors, only that one node have failed the night before but was successfully brought up:
2022-09-26 23:45:13.693 [warning] <0.25764.7> Received a 'DOWN' message from rabbit@yb8162 but still can communicate with it
2022-09-26 23:45:20.827 [info] <0.590.0> rabbit on node rabbit@yb8163 down
2022-09-26 23:45:20.833 [info] <0.590.0> Keeping rabbit@yb8163 listeners: the node is already back
2022-09-26 23:45:20.902 [info] <0.590.0> 8 transient queues from an old incarnation of node rabbit@yb8163 deleted in 0.068931s
2022-09-26 23:45:24.279 [info] <0.590.0> rabbit on node rabbit@yb8163 up
What do we miss here and how to prevent it happening in the future? Is it a known bug and is fixed in the newer vesions?