I have a UCMA based application running as a Trusted App on a Trusted Application Pool (say trust.example.com). For reliability I would like to run it on two servers, trust1.example.com and trust2.example.com, so I plan to add them both servers to the Trusted Application Pool. The application handles IM traffic, so it has a single SIP identity for the application.
My application is active/passive (i.e both instances of the application running on the two servers can not be active at the same time. The passive should kick in only when the active is down.
If a hardware load balancer is not available, and I set up a DNS entry for trust.example.com with 2 IPs (to point to the IP addresses of trust1.example.com and trust2.example.com), I understand that Lync Front End Server (FE) will try to connect to the two IPs on the configured port for my application, in a round robin fashion based on what IP address it gets back from the DNS server.
The question is this: If the FE picks the IP address of trust1.example.com, and my application on trust1.example.com is down, will Lync automatically try to reconnect to trust2.example.com? It has knowledge of the second server because of the Trusted Pool definition.
On failure, will Lync retry to connect to the alternate servers in the Trusted Pool?
Can a DNS based configuration be used in this scenario or will half the traffic to my app get dropped if one of the two servers in the Trusted Application pool goes down?
(The documentation is quiet on this topic and I do not yet have a lab set up to test this)