On Thu, 16 Oct 2008 at 04:46, Antoine wrote:
>  call.cpp file) but did not change anything.  Then I noticed that my
>  RTP library was not installed on the Mandriva client (sorry I am a bit
>  novice to Linux), installed everything related to RTP and still no
>  joy...so I am still working on it...

Well, if the RTP library was actually needed for the RTP transmissions
you'd expect you'd have gotten some sort of error....but I'm talking
here as someone with lots of Linux experience but who is only just a
user of SIPp, who has never tried to do RTP with it.

> -> What would be the maximum number of calls the Windows port of SIPp can 
> -> handle?

I'm sure that will depend on your processor and I/O speed just like
it does on Linux, it's just that the same hardware gets you fewer
calls-per-second under Windows/cygwin.

>  2. I would like to do my testing using a bidirectional calls, I guess
>  the only way would be to run 2 instances of SIPp on both clients, ie
>  one instance for UAS and another one for UAC, ie:

Sounds reasonable to me, but again I've never tried this.  Another
alternative might be IMS Bench SIPp, which I'm currently learning, which
can handle multiple scenarios per SIPp instance, but which is based on
SIPp 1.1.  A second advantage of the IMS Bench is that it has a manager
that can ramp up the CPS rate in steps until too many call failures
(complete call failures or inadequate response time) start happening,
and record the results of each step.  Although...it's oriented toward
putting a SIP proxy or other central service node under load test,
and if you don't have such a node (which it sounds like you don't) then
using the "raw" features of IMS Bench SIPp to set up the kind of tests
you want may be more trouble than it is worth.  It all depends on your
goals and timetable :)

>  Does this sound a sensible way to test my Lan in terms of simultanous
>  bidirectional G711 calls?

Well, speaking as someone who understands something about LANs but is
not an expert, if you are trying to stress test your LAN infrastructure
in the face of VOIP calls, I'd worry that a two point test, while a good
place to start, might not be sufficient.  Having multiple speakers hitting
your switches or whatever would be a more realistic test than just two
speakers going over a single pair of switch ports (unless that's what
your production load is going to look like!).  My thought is that
the two point test might overstate or understate the actual LAN capacity
depending on the exact characteristics of your hardware and software.

--RDM

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Sipp-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sipp-users

Reply via email to