[Asterisk-Dev] channel driver sample - Coding/Transcoding/Conferences

[John Todd]

At 5:40 PM -0500 on 7/4/04, Dr. Rich Murphey wrote:
>  > [mailto:asterisk-dev-admin at lists.digium.com] On Behalf Of John Todd
>>  Even if I
>>  just had seven really cheap ($800) PC's, each with a 4-port
>>  T410P handing their traffic off TDMoE to one DSP-capable
>>  system - that would reduce my costs quite a bit, rather than
>>  buying seven $2800 dual-processor Xeons to get G.729 and
>>  questionable echo cancellation.  Let's say that DSP card cost
>>  me $1500 - I'm still very far ahead.  Heck, I could even buy
>>  TWO complete DSP systems for redundancy and STILL be saving ~$10000.
>
>Where is the bottleneck on a dual xeon with T410P cards without transcoding?
>
>Where is the bottleneck when transcoding is used on all channels?
>
>If two 4-port TDM cards are the recommended maximum, does the interrupt rate
>limit performance, or is it something else?
>
>I hope to be able to test some of this at some point in the near future.
>I'm still gathering the last few parts for a dual xeon to do so.
>
>Rich

Rich -
   I don't know - I wish I did.  This is has fallen off the bottom of 
the list on my "to-do" catalog - I'm so swamped at the moment I can't 
even finish my sent...

   I'd like to see the OSDL (Open Source Development Labs) people pick 
up the ball here for testing, but they've made it clear that they 
only do kernel work and that applications and hardware are outside 
their scope, which is a shame.  It seems pretty meaningful to have 
these data, but there is no group that I can think of that would be 
best suited to perform them.

   Does anyone else have any idea what body would be well-suited to 
test in an unbiased way the capabilities of Asterisk's performance 
(with and without Digium or other interface cards) and then present 
those results to the community?  I suspect we're about to hit the 
wall of "Show me the money" again.

JT