[asterisk-users] Asterisk concurrent calls count

[Jay R. Ashworth]

On Sun, May 18, 2008 at 12:32:01PM -0700, Steve Edwards wrote:
> > I guess I hate to see something I have viewed as such a huge paradigm
> > shift and disruptive force from selling boxes to selling knowledge.
> 
> If I buy a bare DL380 from you, will it cost the same as a fully loaded 
> DL380 configured, optimized, and guaranteed to handle 400 seats? I think 
> you sell your "knowledge" as well as a "box" as well as your time.

This is a good point, so let me expand it:

Steve (Totaro): *this is what your clients are paying you for*.  To
know those answers.

Are you turning around and bitching that someone isn't giving you those
answers for free so you can charge clients for them?  :-)

> In the interest in starting the process, here are a couple of metrics I'd 
> be interested in.
> 
> ) What is the maximum number of simultaneous calls (1 on 1 conversations) 
> that can be bridged before call quality is impaired>
> 
> ) What is the maximum number of simultaneous calls that can be put into a 
> single meetme conference before call quality is impaired>
> 
> ) What is the maximum number of 3 person (agent, customer, supervisor) 
> meetme conferences before call quality is impaired>
> 
> How do you objectively measure call quality? Pass a sine wave at the upper 
> and lower range of a human voice and compare the waveforms?

No, actually there are benches for that; cell providers have a standard
or two, I think.

> How do you construct a "standard" benchmark test bed? 2 identical Asterisk 
> systems, 1 beating on the other?
> 
> These metrics should be run for each technology (IAX, SIP, TDM via T1, TDM 
> via USB) as well as Asterisk version (1.2, 1.4, 1.6)
> 
> There are a lot of variables (OS flavor, OS version, OS tweaks, gcc 
> version, network interface and driver, etc.) that need to be identified 
> and as we collect more samples we may discover that some of these 
> variables are important and some are not. This implies that at least in 
> the "beta" stage of developing a benchmark, submitters must "own" their 
> samples and be willing to re-run tests as the benchmark is refined.

Oh yeah; it's not a small job.

Cheers,
-- jra
-- 
Jay R. Ashworth                   Baylink                      jra at baylink.com
Designer                     The Things I Think                       RFC 2100
Ashworth & Associates     http://baylink.pitas.com                     '87 e24
St Petersburg FL USA      http://photo.imageinc.us             +1 727 647 1274

	     Those who cast the vote decide nothing.
	     Those who count the vote decide everything.
	       -- (Joseph Stalin)