[asterisk-users] Asterisk V/s FreeSwitch

[Bryant Zimmerman]

----------------------------------------
 From: "Patrick Lists" <asterisk-list at puzzled.xs4all.nl>
Sent: Thursday, February 09, 2012 10:42 AM
To: asterisk-users at lists.digium.com
Subject: Re: [asterisk-users] Asterisk V/s FreeSwitch

On 09-02-12 14:52, Stefan Schmidt wrote:
> Am 09.02.12 14:19, schrieb Bryant Zimmerman:
>> Stefan
>>
>> This is on target with my configuration I am working on. What kind of
>> dialplan were you using when running the tests.
>> Were you doing database lookups or just answering the calls and playing
>> hold music. Any example would be appreciated so we can quantify your 
test
>> results. I look forward to your response.
>>
>> Thanks
>> Bryant
>
> the dialplan is quite simple:
>
> for the signaling up to 13500 CC i use this wait and for the 10000CC i
> enable the musiconhold
>
> exten => monitor,1,Noop(PERFORMANCE TESTS)
> exten => monitor,n,Answer
> ;exten => monitor,n,MusicOnHold(806,45)
> exten => monitor,n,Wait(45)
> exten => monitor,n,Hangup

Iirc a long time ago there was a discussion about load testing by 
playing MoH was not a realistic test. Something about all MoH music 
getting streamed synchronized so basically Asterisk only has to stream 
one file and sorta multiplex that single output to all the established 
calls (legs).

[snip]

> btw my normal production machines which are just the same virtual
> machines like this test system. i also had 330 concurrent calls, some
> with transcoding, many database lookups, musiconhold, pickup ... and the
> sysload was around 1.0 ;)

The difference (13500 with MoH versus 330 with a real dialplan) shows 
that it makes sense to mimic your dialplan in your test scenario as much 
as possible to see how far you can realistically push the box and still 
keep things stable and sound quality good.

Regards,
Patrick

-------------------------------------
Patrick

I agree with you but it looks like these test show that asterisk 10 could 
handle a very high volume switch style application if you remove the rtp 
and media handling from the dialplan. For me this opens up the question as 
to why would I need freeswitch for high volume switching if my RTP is being 
handled else where. I would have a different dialplan code set for this 
kind of application anyway. This is why I asked Stefan to share his 
dialplan and testing matrix. It allows us to see raw performance what is 
the asterisk code base cable of and for his test case it is quite 
impressive.  For other test cases it may not be. I would like to see 
someone test a full PBX implementation with independent audio files playing 
not just MOH and see what kind of load it can handle there. That would give 
me a better idea of how the two would compare. I wounder how the new 10.0 
conference application would stack against freeswitches conference. How 
would we best design a test for that?

This is a great conversation. Very productive for me at least. 

Thanks
Bryant
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-users/attachments/20120209/fd44de4b/attachment.htm>