Has any seen tunnels bounce a lot between the Nod and the relay hubs?
I have 4 relay hubs and about 70 client hubs. The queues for the client hubs terminate at the one of the relay hubs.
Have not hit the queue/subscriber limitation and have upgraded the hubs to the latest version 5.66.
Because one of the symptoms also seen are queues backing up, I would think the bulk sizes need to be adjusted for the relay queues and the data engine to allow messages to flow through faster.
Any ideas out there?