I tried this with two A records for SBC1 pointing to two IPs and no luck. I had to bounce the services to get any DNS changes to kick in but it didn't ever try to connect to the second SBC IP address. I determined this from a Wireshark trace and no SYN requests heading out so I don't think it even tries the second DNC entry.
One good thing is I could see that the failover works well from SE1 server to SE2 server at dial time. If SE1 gateway is down it tries for 5-10 seconds and then moves onto the SE2 gateway whilst in the same call, so no need to redial. Possibly on the downside you get a ring tone even when it can't get a TCP connection to the SBC, so it hides the failyre to the end user which prob makes debugging a little harder.
So the testing I have performed so far doesn't help me work out what happens when the session limit is reached for a SIP trunk ... anybody have a clue on this one?
Jed
