Is there a way to check the status of rabbitmq node data synchronization ?

[xixiangzouyibian]

For example:
We deployed 3 rabbitmq nodes, each rabbitmq node deployed on one k8s node (anti-affinity).
Now one of rabbitmq node broken for some reasons, we migrate the broken rabbitmq node to idle k8s node then we get a healthy three-node rabbitmq cluster.
But from my understanding, the new deployed rabbitmq node will take the time to do the fully data synchronization from master rabbitmq node. so how do I know if data sync is finish ?
It seems query rabbitmqcluster may not work:

NAME ALLREPLICASREADY RECONCILESUCCESS AGE
rabbitmq True True 4d

But elasticsearch provides the field(HEALTH) to query:

NAME HEALTH NODES VERSION PHASE AGE
catalog green 3 7.17.10 Ready 4d
search green 3 7.17.10 Ready 4d

[xixiangzouyibian]

Any updates ?

[lukebakken]

It is impolite to bump your own question. It has only been four days, be patient.

[Zerpet]

There's no practical way to verify this. Most of the time you don't need to worry about this. RabbitMQ will do it's replication and it will complete relatively fast. If you want to inspect a specific queue, you have the command rabbitmq-queues quorum_status or rabbitmq-streams stream_status

[xixiangzouyibian]

@Zerpet Thanks for you response.
Is there any data that can indicate that the data synchronization between nodes is fast ? Could you evaluate the magnitude of the synchronization time, seconds level or millisec level ? And will the synchronization time become longer when there is more data, more queues or more rabbitmq nodes.

[michaelklishin]

Infrastructure-level metrics can.

[michaelklishin]

Starting with RabbitMQ 3.13, new quorum queue replicas will not be considered "joined" (online, counting towards the quorum) if they are performing their initial sync. All subsequent syncs should transfer drastically less data (well, proportionally to how far behind this replica is, of course).

[xixiangzouyibian]

Thanks Michael.