papa-bear

I can see redundancy in both gateways and signaling servers being a benefit but if your site is spread out like ours, try and get diverse routes of fibre (or cable) into a site with multiple paths back to where the core equipment. That is where the backhoe analogy kicks in.

I am thinking of multiple gateways for both incoming and outgoing calls. The incoming would most likely be split up and if one site went down a % of people would loose incoming calls (unless there is an inexpensive way to have the telcos re-route DIDs and outgoing calls would almost always have service (provided there are multiple paths to the sites). Put survivable gateways with a single 1B into smaller sites incase there is a failure, they will at least have dial tone in an emergency (plus it can be used as an identifying line for 911 calls if they are caught in the gateway and routed out).

There are so many ways to mitigate the risks. One almost needs to know how much $$ they have available then model it, present it and if there is not enough availablility in the system then point out how much more they will need to spend.