Michael Loftis wrote:
It seems that disabling MMX support in zaptel fixed *all* my problems,
from hold music, to iLBC times, to random crashes. That's odd though, as
I'm using a Celeron 1.7 GHz chip which supports MMX. Perhaps the small
celeron cache is to blame?
Could be, celeron's are also kn
--On Friday, October 29, 2004 11:08 -0700 Trevor Peirce
<[EMAIL PROTECTED]> wrote:
Michael Loftis wrote:
Little tireds now so you may have already done all this but make sure
you have latest libpri, zaptel, and asterisk, in that order.
It seems that disabling MMX support in zaptel fixed *all* m
Michael Loftis wrote:
Little tireds now so you may have already done all this but make sure
you have latest libpri, zaptel, and asterisk, in that order.
It seems that disabling MMX support in zaptel fixed *all* my problems,
from hold music, to iLBC times, to random crashes. That's odd though,
Senad Jordanovic wrote:
>
> Trevor Peirce wrote:
>
> Trevor,
>
> You are better off using this instead of mpg321:
>
> http://bugs.digium.com/bug_view_page.php?bug_id=0002379
>
> Mark... Any chance including this patch in the 1.0?
NO! 1.0 is a *STABLE* branch. *ONLY* bug fixes should go into
Michael Loftis wrote:
The fact that it crashes says something more is wrong naywa...have you
tried running anything like Memtest86 on the box fora time just to
see if it comes up with anything?
Yes I have allowed memtest86 to run for about 9 hours overnight... that
is one of the first things I
The fact that it crashes says something more is wrong naywa...have you
tried running anything like Memtest86 on the box fora time just to see if
it comes up with anything?
Little tireds now so you may have already done all this but make sure you
have latest libpri, zaptel, and asterisk, in tha
Okay, I have removed the IDE Controller and am now using onboard. The
problems below still exist--
Trevor Peirce wrote:
show translation still reveals the iLBC column in the 8700 to 9600
range though. LPC10's row is also in the 900s.
show translation recalc 10 still causes the * console to st
On October 23, 2004 10:58 pm, Michael Loftis wrote:
> mmm... any packaging is better than none. I regularly destroy things on
> systems when it's not been put into proper packaging because we upgrade the
> system, and there's no record of something being installed, nor what it
> depends on, so it
Todd Lieberman wrote:
Wo trevor, Format and start over? Don't go crazy, just remove the files
created by make install.
Fighting for weeks to get a more-or-less stable telephone system can
drive a man to do extraordinary things like rebuilding a server from
scratch!
We are making process, howeve
--On Saturday, October 23, 2004 21:35 -0500 Brian West <[EMAIL PROTECTED]>
wrote:
Done and done. FYI you may want to update
http://voip-info.org/wiki-Asterisk+mpg123+redhat as that's where I got
inspired to download the RPM.
Repeat after me... RPM is bad source is good.
I have put a nice wa
> Done and done. FYI you may want to update
> http://voip-info.org/wiki-Asterisk+mpg123+redhat as that's where I got
> inspired to download the RPM.
Repeat after me... RPM is bad source is good.
I have put a nice warning on that page. Its already been proven to use
0.59r and you'll notice t
ist - Non-Commercial Discussion
> Subject: Re: [Asterisk-Users] iLBC/PCM16 Huge Cost
>
> Brian West wrote:
>
> >REMOVE THAT POS and install mpg123 0.59r, compile from src.
> >
> Done and done. FYI you may want to update
> http://voip-info.org/wiki-Asterisk+mpg123+redha
rcial Discussion
Subject: Re: [Asterisk-Users] iLBC/PCM16 Huge Cost
Brian West wrote:
>REMOVE THAT POS and install mpg123 0.59r, compile from src.
>
Done and done. FYI you may want to update
http://voip-info.org/wiki-Asterisk+mpg123+redhat as that's where I got
inspired to download the RPM.
Trevor Peirce wrote:
> Trevor Peirce wrote:
>
>> I have noticed that when * is first loading, CPU usage goes to 100%
>> for exactly the same duration that it takes that ilbc codec to load.
>
>
> Upon closer inspection, it seems that every time a caller is hears
> MOH, a new mpg123 is spawned. R
Brian West wrote:
REMOVE THAT POS and install mpg123 0.59r, compile from src.
Done and done. FYI you may want to update
http://voip-info.org/wiki-Asterisk+mpg123+redhat as that's where I got
inspired to download the RPM.
I just stopped asterisk and killed off all the mpg123 processes... ran
sa
cial Discussion
> Subject: Re: [Asterisk-Users] iLBC/PCM16 Huge Cost
>
> Brian West wrote:
>
> >>Upon closer inspection, it seems that every time a caller is hears MOH,
> >>a new mpg123 is spawned. Right now top is showing 8 mpg123's running,
> >>and
Brian West wrote:
Upon closer inspection, it seems that every time a caller is hears MOH,
a new mpg123 is spawned. Right now top is showing 8 mpg123's running,
and between then and * CPU utilisation is maxed out. About 30% user and
65% system.
IMPOSSIBLE... What mpg123 version are you runni
> Upon closer inspection, it seems that every time a caller is hears MOH,
> a new mpg123 is spawned. Right now top is showing 8 mpg123's running,
> and between then and * CPU utilisation is maxed out. About 30% user and
> 65% system.
IMPOSSIBLE... What mpg123 version are you running?
bkw
___
Trevor Peirce wrote:
I have noticed that when * is first loading, CPU usage goes to 100%
for exactly the same duration that it takes that ilbc codec to load.
Upon closer inspection, it seems that every time a caller is hears MOH,
a new mpg123 is spawned. Right now top is showing 8 mpg123's runn
Eric Wieling wrote:
Kevin Walsh wrote:
Benjamin on Asterisk Mailing Lists [EMAIL PROTECTED]
wrote:
Can you set up a test call where Asterisk will transcode from ulaw to
ILBC and see what it does to your CPU load?
How should I go about creating such a test call?
Also, try recalculating the transl
Kevin Walsh wrote:
Benjamin on Asterisk Mailing Lists [EMAIL PROTECTED] wrote:
On Sat, 23 Oct 2004 02:16:44 -0700, Trevor Peirce <[EMAIL PROTECTED]> wrote:
G723 GSM ULAW ALAW G726 ADPCM SLINR LPC10 G729A SPEEX ILBC
G723 - - - - - - - - - -
ling List - Non-Commercial Discussion
> Subject: Re: [Asterisk-Users] iLBC/PCM16 Huge Cost
>
> On Sat, 23 Oct 2004 02:16:44 -0700, Trevor Peirce <[EMAIL PROTECTED]>
> wrote:
> > G723 GSM ULAW ALAW G726 ADPCM S
Benjamin on Asterisk Mailing Lists [EMAIL PROTECTED] wrote:
> On Sat, 23 Oct 2004 02:16:44 -0700, Trevor Peirce <[EMAIL PROTECTED]> wrote:
> > G723 GSM ULAW ALAW G726 ADPCM SLINR LPC10 G729A SPEEX ILBC
> > G723 - - - - - - - - - - -
> >
On Sat, 23 Oct 2004 02:16:44 -0700, Trevor Peirce <[EMAIL PROTECTED]> wrote:
> G723 GSM ULAW ALAW G726 ADPCM SLINR LPC10 G729A SPEEX ILBC
> G723 - - - - - - - - - - -
>GSM - - 2 2 4 2 1 1238 - - 529
Could you tell us what RAID card you are using + what drivers you are using
for it.
Could you try to run it without the raid card ?
Zoa.
At 12:35 23/10/2004, you wrote:
Trevor Peirce wrote:
Sure. Running Fedora Core 2 with latest updates (kernel 2.6.8-1.521) on
a Celeron 1.70 GHz chip. Half a
Trevor Peirce wrote:
Sure. Running Fedora Core 2 with latest updates (kernel 2.6.8-1.521) on
a Celeron 1.70 GHz chip. Half a gig DDR ram, one generic X100P card
with it's very own IRQ.
Asterisk is the latest CVS. It's about time for bed.. spent too many
hours trying to figure out other things t
joachim wrote:
Could you give us more information on:
Distro, kernel version, compiler, makefile flags, version of asterisk,
and hardware on your machine, + loaded modules ?
GSM to LPC10 is also way tooo slow.
Sure. Running Fedora Core 2 with latest updates (kernel 2.6.8-1.521) on
a Celeron 1.70
Could you give us more information on:
Distro, kernel version, compiler, makefile flags, version of asterisk, and
hardware on your machine, + loaded modules ?
GSM to LPC10 is also way tooo slow.
-
*CLI> show uptime
System uptime: 27 minutes, 2 seconds
*CLI> show translation
Tran
joachim wrote:
I have seen similar things in the past, but only during startup.
When started, do a show translation and look again, if that value is
ok, you can ignore the one on startup.
*CLI> show uptime
System uptime: 27 minutes, 2 seconds
*CLI> show translation
Translation times between
I have seen similar things in the past, but only during startup.
When started, do a show translation and look again, if that value is ok,
you can ignore the one on startup.
Zoa.
At 12:06 23/10/2004, you wrote:
Hello,
During asterisk bootup, I've been having a fun time with a random delay
which
Hello,
During asterisk bootup, I've been having a fun time with a random delay
which can be quite long, from what seems to be the codec_ilbc.so file.
I notice in verbose mode the cost is rather high, and was hoping someone
will have some insight on what's going on here.
Prior to a harddrive dyi
31 matches
Mail list logo
