The only true voice switch...
Is a class 5 switch. When they get overwhelmed, they don't crash; you just get a busy or fast busy (re-order) tone until a facility is free. But they cost money, especially with less competition (we miss you Nortel).
The switch wasn't the problem. The problem was that the number of inbound calls exceeded the number of available trunks. It's likely that OnNet calls and 911 calls were not effected.
As mentioned in my previous post, when a class 5 switch experiences excessive inbound or outbound non-local traffic, the person placing the call will receive either a busy tone or a reorder tone. But as individual trunks come free, others can access it. Just keep trying. As for local traffic, that would remain unaffected as it is nearly impossible to have more locally switched calls connected then time slots available. Even with an equipment malfunction, redundancy prevents the customer from being affected.
The symptoms described by forum posters indicate there was no service at all. This would seem to imply some sort of system wide 'crash', meaning no service at all to the affected customers.
I am just trying to point out that you get what you pay for.
I hate to break it to you but my softswitch would do the exact same thing that your beloved 5ESS would do. The article didn't give any indication that there was a system wide crash. Only that inbound and outbound calls were effected.
Most people won't hear a reorder anyways. Most (maybe all) of the cell providers substitute a reorder with a "You have reached a non working number" message.
A little secret there is not a one to one switch connection it is about 70%. The same with telco trunks out of and in to an office. The same thing happens when every kid picks up the phone on a snow day only 7 out of 10 get dial tone. Other then helping with price by competition you got what you paid for. It will not make much of a difference since all dial tone will be VoIP when everything is just data over ethernet.
|reply to battleop |
I'd be just as concerned about the spike in bandwidth and cpu cycles being eaten up at the CMTS due to this ridiculous amount of stress on the NCS system. Usually calls barely register as activity at the CMTS but during peak bandwidth times it might pose an issue.