I finally took the plunge and going full corona, updating the farm and pipeline right now.
One issue I'd like to get some advise on is the following, I've got 2 farms, a small one in my studio (3pc) and the heavy projects I renders on EC2 where I've got a nice scalable setup in a VPC which is connected to my LAN via a VPN.
the problem is that there isn't a single machine that all these slaves can reach.. my workstation is the lic server for the in-house farm which work fine, and the 'manager' node on the cloud is dealing with the lics for the cloud instances. (the cloud slaves can only see the manager node). Besides that the cloud render slaves come and go (spot), and the the manager is a dedicated instance. (see attached sketch)
So my questions are:
-Does running 2 lic servers work fine or should I find a solution where there is only one such server
-If a cloud slave goes down is that lic immediate available again for a new node? (I don't see any notification about lics being free'ed up in the logs)
Any other issues I might run into?