[walkman]

For those of you still on-board with this thread, I was able to put in a workaround. I had the customer open the fixed ports on the PIX and added static translations for them along with rules to allow them. This enables the call to continue indefinitely since the H.245 traffic now uses a busy RTP port on the PIX.

Interesting side note, once the connection starts using the fixed ports the "RoundTripDelayReq" and "RoundTripDelayResp" packets were evident every 30 seconds just like with the other endpoint/bridge combinations. Not sure why, but seems that the PIX was blocking that stuff along with the keepalives (as suggested by Joe V).

I still have some investigation to complete, but the workaround is in place for now and the customer is happy.

I very much appreciate the help I got on this board. The level of expertise is comforting! I'm sure I'll be back for more. Thanks all!