[asterisk-biz] Does VoIP Really Work for Serious Business?

[Herman Webley]

On Tue, 2006-03-07 at 15:21 -0500, Steve Totaro wrote:
> This is a great and honest email.  I am looking forward to the replies.
> 
> My thoughts are VoIP is ready for serious business on the LAN and point
> to point links.

I think VoIP, open source VoIP and Asterisk have a lot to offer.

Despite all the claims that TDM is the best, VoIP offers alot that is
difficult or simply too expensive to implement with conventional
telephony.

Small office
------------
Open source VoIP will have a hard time competing here. Ultra low cost
embedded devices will take this one. They'll use VoIP though, and we can
count that as a victory (we can go in and add features as the business
grows). 

Enterprise/SMB
--------------
  On a private network VoIP beats TDM hands down. In properly QoS WAN
scenarios (leased lines, MPLS, FR etc) VoIP still beats TDM. Here are
some random advantages that I see for these two implementation models:

1.cheaper hardware and software
2.can use wideband codecs (better sound quality)
3.can use video
4.can provide backup servers, redundant paths, etc. more cheaply/easily
(sure you can by a nifty TDM PBX with all sorts of redundancy, but say
you want to double whatever redundancy it provides, can you buy more
commodity PCs, switches, cat5 and let linux HA+STP do the rest? Or do
you have to give your arm, leg and 1st born?)

Hence, one can engineer voice systems for (a)better sound, (b)video and
(c) greater reliability. Why would we want to spent more money to go
with obsolete technology that won't deliver those features?

Open source VoIP adds arbitrarily cool features implemented in software
(who writes cool modules for IOS?).

Service provider
----------------
  Here the network is again controllable. VoIP works great. VoIP saves
loads of money over conventional technology solutions (even if you go
non-open source like cisco or such). Using Asterisk as a feature server
in a service provider scenario rocks. The only thing I can think of
right now that is missing is maintaining calls through server failover.
If that is added we can make the system as fault tolerant and redundant
as we wish.

As far as open source VoIP in concerned, a lot needs to be done for
gateways to legacy systems and infrasture. Wheres the channelized DS3
interface? And what would we do with that anyway? Where is the tried,
true, and certified SS7 solution (I want more than F links damnnit :P).
[Why can't a open source solution be certified? Why do people believe
that just because a solution is 'commercial' that it has been tested and
certified?]
Where are the DS3 channels banks? I want to provision copper loops
damnit :P How do I add 20,000 lines to an existing telco that wants to
use VoIP? Do I buy a thousand 'high density' SIP gateways?


Public internet sucks, but that is nothing new and it dosn't just suck
for VoIP. I want my telecom provider to give me VoIP trunks over a line
they run to my premises.

Cheers,
Herman