Few weeks ago I wrote about the first voice portals, back in ’97.
Fast forward 5 years. VoIP communication is all the rage and we all merrily jump in the wagon (we do love us a new technology, don’t we) and start working on the network. All without knowing a thing about IP stuff, hoping that the behavior of our nice little voice toys, on IP transport, would be the same. And obviously it is not. I remember getting a call from a customer one afternoon timidly complaining that their voice responses (and in fact the whole contact center) was not answering calls: it was, in fact, dropping calls. Now, dropping calls in a telephony environment is like the worst thing that could happen, only second to a complete major outage of the platform (or death in the family), so I righteously panicked and called the network guy that was in charge of the VoIP configuration. He answered calmly that he got the customer call as well and that he did get immediately into the system, checked his beloved router and saw that all was fine, so he scheduled the support for a more convenient day. He would look at the issue the next day as soon as he was in the office.
Well, that was my first encounter with the “network guys”, where a bit of network outage did not mean _panic and destruction_ but only a few packets lost here and there, no fuss, no big deal. Hence the first lesson learned with VoIP: synchronous communication does mind if packets are lost, and even a few seconds outage IS a big deal and will probably at best miss portions of the interaction, at worst drop the whole connection. In the end it was the carrier’s fault and the connection with the voice gateway took all the blame, but after this occurrence the network guy never missed a “dropped calls” alert again. 🙂