[Asterisk-Dev] Is there a need for a DS3 channel/driver?

[Clint Guillot]

On Nov 18, 2004, at 4:17 PM, Race Vanderdecken wrote:
> Let us stipulate for the moment that it will work, given a fast server
> to run it on.
>
> What is the business use for such a toy?
>
A DS3 is quite a bit cheaper than umpteen t-1's.

We are evaluating Asterisk using 4 T1's into a digium card.  If 
(probably) this solution becomes the replacement production system, we 
will require a lot more channels to handle the load, so the next step 
in our evaluation it throwing 4x 4 port digium cards in a chassis to 
see how it keeps up.
If this works out, boxes with 16 t1's in them will be the production 
platform.  Not a huge leap from there to a DS3 if the hardware can 
handle double the data...  And DS3's are cheaper than 16 T1's, which 
gives us some 'free capacity.'

So, to answer your question more fully, the business case for a DS3 
card is:

Increased density and the attendant decrease in maintenance and 
configuration concerns
Recurring transport charge savings (many t1's vs 1 ds3)
Equipment (DS-3 channel bank) savings (not hugely significant 
cost-wise, but it's one more thing to maintain)

Down Side:

Single point can cause more widespread failure (can be overcome by 
scale)
Unknown requirements for host processor may be prohibitive (technically 
or economically)
Many fewer potential installations than T1 based equipment means less 
thoroughly tested and fewer folks to help fix issues

My conclusion:
Though a DS3 card is not a *requirement* for Asterisk currently, like 
SS7, it will become a requirement long term.  In the mean time, it will 
be a nicety for some.

Let the dissection begin :)

Clint