Hi.I need to know aproximately, how many H323 calls coming from a ethernet 10/100 card , i can manage with Asterisk running on a Pentium III 750 MHZ machine with Linux RH9.Any source of bechmarks ? Thank you. Rodrigo.
---------------------------------
Internet gratis: usuario "yahoo", contraseña "yahoo". Desde Buenos Aires, 4004-1010. Más ciudades
--0-1572319603-1050521989=:44522
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
<DIV>Hi.</DIV>
<DIV>I need to know aproximately, how many H323 calls coming from a ethernet 10/100 card , i can manage with Asterisk running on a Pentium III 750 MHZ machine with Linux RH9.</DIV>
<DIV>Any source of bechmarks ?</DIV>
<DIV> </DIV>
<DIV>Thank you.</DIV>
<DIV> </DIV>
<DIV>Rodrigo.</DIV><p><br><hr size=1><b><a href="http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com">Internet gratis</a>: usuario "yahoo", contraseña "yahoo". Desde Buenos Aires, 4004-1010. <a href="http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com/avanzados.html">Más ciudades</a></b>
--0-1572319603-1050521989=:44522--
This is a multi-part message in MIME format.
--------------070600000408040201050003
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Unfortunately, it depends on which channel driver your running.
chan_h323 is setup so that you can take advantage of the codec
pass-through feature of Asterisk. We have simulated 125 concurrent
calls setup in random (sub-15 second) intervals, using passthru, on a
500mhz P4 without causing more than a 5 second delay in call setup
time. We have a few separate production systems (single 2.53 ghz, P4,
2gig ram) with running pass-thru and have never had a single compliant
about audio problems, but at this point we've only pushed about 3 T-1s
of concurrent real traffic on any one system.
Now, if you have to do the actual codec-ing on the machine then you are
going to lose overall number of calls due to the processor time it takes
to encode and decode whatever codec your running. On that same 500mhz
P4 test machine, we were able to simulate 45 concurrent calls using GSM,
again without a long delay in call setup time.
With that other H.323 driver we couldn't make10 concurrent calls work
reliably, so we wrote chan_h323.
Jeremy McNamara
Xara Developement wrote:
Quote:
Hi.
I need to know aproximately, how many H323 calls coming from a
ethernet 10/100 card , i can manage with Asterisk running on a Pentium
III 750 MHZ machine with Linux RH9.
Any source of bechmarks ?
Thank you.
Rodrigo.
------------------------------------------------------------------------
*Internet gratis
<http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com>: usuario
"yahoo", contraseña "yahoo". Desde Buenos Aires, 4004-1010. Más
ciudades
<http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com/avanzados.html>*
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<title></title>
</head>
<body>
<br>
Unfortunately, it depends on which channel driver your running.
chan_h323 is setup so that you can take advantage of the codec
pass-through feature of Asterisk. We have simulated 125 concurrent
calls setup in random (sub-15 second) intervals, using passthru, on a
500mhz P4 without causing more than a 5 second delay in call setup
time. We have a few separate production systems (single 2.53 ghz, P4,
2gig ram) with running pass-thru and have never had a single compliant
about audio problems, but at this point we've only pushed about 3 T-1s
of concurrent real traffic on any one system.<br>
<br>
Now, if you have to do the actual codec-ing on the machine then you are
going to lose overall number of calls due to the processor time it
takes to encode and decode whatever codec your running. On that same
500mhz P4 test machine, we were able to simulate 45 concurrent calls
using GSM, again without a long delay in call setup time.<br>
<br>
With that other H.323 driver we couldn't make10 concurrent calls work
reliably, so we wrote chan_h323.<br>
<br>
<br>
Jeremy McNamara<br>
<br>
<br>
Xara Developement wrote:<br>
<blockquote type="cite"
cite="mid20030416193949.45124.qmail@web41804.mail.yahoo.com">
<div>Hi.</div>
<div>I need to know aproximately, how many H323 calls coming from a
ethernet 10/100 card , i can manage with Asterisk running on a Pentium
III 750 MHZ machine with Linux RH9.</div>
<div>Any source of bechmarks ?</div>
<div> </div>
<div>Thank you.</div>
<div> </div>
<div>Rodrigo.</div>
<p><br>
</p>
<hr size="1"><b><a
href="http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com">Internet
gratis</a>: usuario "yahoo", contraseña "yahoo". Desde Buenos Aires,
4004-1010. <a
href="http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com/avanzados.html">Más
ciudades</a></b> </blockquote>
<br>
</body>
</html>
Hello,
Has anyone done any Sip simulations ?. I need to know how many
concurrent Sip calls that can be handled with and without codec-ing.
Regards,
Abdul Hakeem
-----Original Message-----
From: asterisk-dev-admin@lists.digium.com
[mailto:asterisk-dev-admin@lists.digium.com] On Behalf Of Jeremy
McNamara
Sent: Wednesday, April 16, 2003 9:52 PM
To: asterisk-dev@lists.digium.com
Subject: Re: [Asterisk-Dev] H323 processing limit
Unfortunately, it depends on which channel driver your running.
chan_h323 is setup so that you can take advantage of the codec
pass-through feature of Asterisk. We have simulated 125 concurrent
calls setup in random (sub-15 second) intervals, using passthru, on a
500mhz P4 without causing more than a 5 second delay in call setup time.
We have a few separate production systems (single 2.53 ghz, P4, 2gig
ram) with running pass-thru and have never had a single compliant about
audio problems, but at this point we've only pushed about 3 T-1s of
concurrent real traffic on any one system.
Now, if you have to do the actual codec-ing on the machine then you are
going to lose overall number of calls due to the processor time it takes
to encode and decode whatever codec your running. On that same 500mhz
P4 test machine, we were able to simulate 45 concurrent calls using GSM,
again without a long delay in call setup time.
With that other H.323 driver we couldn't make10 concurrent calls work
reliably, so we wrote chan_h323.
Jeremy McNamara
Xara Developement wrote:
Hi.
I need to know aproximately, how many H323 calls coming from a ethernet
10/100 card , i can manage with Asterisk running on a Pentium III 750
MHZ machine with Linux RH9.
Any source of bechmarks ?
=20
Thank you.
=20
Rodrigo.
_____ =20
Internet <http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com>
gratis: usuario "yahoo", contrase=F1a "yahoo". Desde Buenos Aires,
4004-1010. M=E1s
<http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com/avanzados.h
tml> ciudades=20
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">
<TITLE>Message</TITLE>
<META content=3D"MSHTML 6.00.2600.0" name=3DGENERATOR></HEAD>
<BODY>
<DIV><SPAN class=3D471233721-16042003><FONT=20
face=3D"Book Antiqua">Hello,</FONT></SPAN></DIV>
<DIV><SPAN class=3D471233721-16042003><FONT face=3D"Book Antiqua">Has =
anyone done=20
any Sip simulations ?. I need to know how many concurrent Sip calls that =
can be=20
handled with and without codec-ing.</FONT></SPAN></DIV>
<DIV><SPAN class=3D471233721-16042003><FONT=20
face=3D"Book Antiqua">Regards,</FONT></SPAN></DIV>
<DIV><SPAN class=3D471233721-16042003><FONT face=3D"Book Antiqua">Abdul=20
Hakeem</FONT></SPAN></DIV>
<DIV></DIV>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><FONT face=3DTahoma=20
size=3D2>-----Original Message-----<BR><B>From:</B>=20
asterisk-dev-admin@lists.digium.com =
[mailto:asterisk-dev-admin@lists.digium.com]=20
<B>On Behalf Of </B>Jeremy McNamara<BR><B>Sent:</B> Wednesday, April 16, =
2003=20
9:52 PM<BR><B>To:</B> asterisk-dev@lists.digium.com<BR><B>Subject:</B> =
Re:=20
[Asterisk-Dev] H323 processing =
limit<BR><BR></FONT></DIV><BR>Unfortunately, it=20
depends on which channel driver your running. chan_h323 is setup =
so that=20
you can take advantage of the codec pass-through feature of=20
Asterisk. We have simulated 125 concurrent calls setup in =
random=20
(sub-15 second) intervals, using passthru, on a 500mhz P4 without =
causing=20
more than a 5 second delay in call setup time. We have a few =
separate=20
production systems (single 2.53 ghz, P4, 2gig ram) with running =
pass-thru=20
and have never had a single compliant about audio problems, but at this =
point=20
we've only pushed about 3 T-1s of concurrent real traffic on any one=20
system.<BR><BR>Now, if you have to do the actual codec-ing on the =
machine then=20
you are going to lose overall number of calls due to the processor time =
it takes=20
to encode and decode whatever codec your running. On that =
same=20
500mhz P4 test machine, we were able to simulate 45 concurrent calls =
using GSM,=20
again without a long delay in call setup time.<BR><BR>With that other =
H.323=20
driver we couldn't make10 concurrent calls work reliably, so we wrote=20
chan_h323.<BR><BR><BR>Jeremy McNamara<BR><BR><BR>Xara Developement =
wrote:<BR>
<BLOCKQUOTE cite=3Dmid20030416193949.45124.qmail@web41804.mail.yahoo.com =
type=3D"cite">
<DIV>Hi.</DIV>
<DIV>I need to know aproximately, how many H323 calls coming from a =
ethernet=20
10/100 card , i can manage with Asterisk running on a Pentium III 750 =
MHZ=20
machine with Linux RH9.</DIV>
<DIV>Any source of bechmarks ?</DIV>
<DIV> </DIV>
<DIV>Thank you.</DIV>
<DIV> </DIV>
<DIV>Rodrigo.</DIV>
<P><BR></P>
<HR SIZE=3D1>
<B><A=20
=
href=3D"http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com">Inte=
rnet=20
gratis</A>: usuario "yahoo", contrase=F1a "yahoo". Desde Buenos Aires, =
Posted: Thu Apr 17, 2003 10:32 am Post subject: [Asterisk-Dev] H323 processing limit
Xara Developement wrote:
Quote:
Hi.
I need to know aproximately, how many H323 calls coming from a ethernet
10/100 card , i can manage with Asterisk running on a Pentium III 750
MHZ machine with Linux RH9.
Any source of bechmarks ?
Using asterisk-oh323 we are able to handle ~30 concurrent H.323 calls
on a Celeron 600MHz with 320MB RAM without problems. More than half of
these are pass-through and on the rest some transcoding is
done (GSM <-> G.711).
With the same software on an embedded StrongARM based platform with
64MB RAM we can handle ~10 concurrent H.323 calls, all pass-through.
Michael Manousos.
Quote:
Thank you.
Rodrigo.
------------------------------------------------------------------------
*Internet gratis
<http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com>: usuario
"yahoo", contraseña "yahoo". Desde Buenos Aires, 4004-1010. Más ciudades
<http://ar.rd.yahoo.com/mail/tag/?http://ar.online.yahoo.com/avanzados.html>*
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum