> No but I'm still interested in solving the problem. Right now we are
> using ubuntu-server for this without issue.
I hear you...
Quote:
this will load ztdummy for timing
It's confirmed to be loaded. The problem is that when the box has any
other load to deal with, my zttest accuracy goes down considerably
(low power machine, effects are exaggerated).
-Michael I.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
On Wednesday 25 July 2007 12:17:51 pm Michael Iedema wrote:
Quote:
> > No but I'm still interested in solving the problem. Right now we are
> > using ubuntu-server for this without issue.
I hear you...
> this will load ztdummy for timing
It's confirmed to be loaded. The problem is that when the box has any
other load to deal with, my zttest accuracy goes down considerably
(low power machine, effects are exaggerated).
-Michael I.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
you might want to also look at app conference. I have 1.11. it does not
require timing.
On Wednesday 25 July 2007 12:17:51 pm Michael Iedema wrote:
Quote:
> > No but I'm still interested in solving the problem. Right now we are
> > using ubuntu-server for this without issue.
I hear you...
> this will load ztdummy for timing
It's confirmed to be loaded. The problem is that when the box has any
other load to deal with, my zttest accuracy goes down considerably
(low power machine, effects are exaggerated).
-Michael I.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
On Jul 25, 2007, at 3:17 PM, Michael Iedema wrote:
Quote:
> this will load ztdummy for timing
It's confirmed to be loaded. The problem is that when the box has any
other load to deal with, my zttest accuracy goes down considerably
(low power machine, effects are exaggerated).
I'm having the same issue and checked that my clock is at the FreeBSD
defaults, but I have a TDM04B card installed. Would I still need the
ztdummy?
Cal
--
S i x F e e t U p , I n c . | "Nowhere to go but open source"
Silicon Valley: +1 (650) 401-8579 x602
Midwest: +1 (317) 861-5948 x602
Toll-Free: 1-866-SIX-FEET
mailto:calvin@sixfeetup.com
http://www.sixfeetup.com | Zope/Plone Custom Development
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
On Thursday 26 July 2007 01:33:48 pm Calvin Hendryx-Parker wrote:
Quote:
On Jul 25, 2007, at 3:17 PM, Michael Iedema wrote:
>> this will load ztdummy for timing
>
> It's confirmed to be loaded. The problem is that when the box has any
> other load to deal with, my zttest accuracy goes down considerably
> (low power machine, effects are exaggerated).
I'm having the same issue and checked that my clock is at the FreeBSD
defaults, but I have a TDM04B card installed. Would I still need the
ztdummy?
Cal
I would say yes. as that I have not seen timing work right with a few cards
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
I'm attempting to get my head around the ztdummy code and have some
potentially stupid questions.
How can zttest judge the accuracy of ztdummy if it is a userland
process? Why not use its timing mechanism in ztdummy? (yes, that's a
stupid question...I just want some more info)
Is the accuracy of ztdummy ultimately dependent on timeout() in
kern_timeout.c or can it be affected by the callout passed to it?
Are there any other commonly available timing sources in the FreeBSD
kernel? Is this
(http://www.freebsd.org/cgi/cvsweb.cgi/ports/emulators/rtc) of
interest?
Lastly and most importantly:
- Are any others seeing this drift?
- Can ztdummy be expected (regardless of OS) to be accurate under heavy load?
This is all just thinking out loud, but perhaps it will get some
conversation going as to what the issue really is with ztdummy.
Enjoy your weekend, everyone!
-Michael I.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
Posted: Sat Jul 28, 2007 12:09 am Post subject: [Asterisk-bsd] MeetMe drift
(Oops, I messed up my From: address on that last email...trying again)
Quote:
Hello,
I'm attempting to get my head around the ztdummy code and have some
potentially stupid questions.
How can zttest judge the accuracy of ztdummy if it is a userland
process? Why not use its timing mechanism in ztdummy? (yes, that's a
stupid question...I just want some more info)
Is the accuracy of ztdummy ultimately dependent on timeout() in
kern_timeout.c or can it be affected by the callout passed to it?
As you've noticed, ztdummy uses FreeBSD's timeout() mechanism to schedule
its interrupts. timeout() is timed using the interval clock, typically
1000Hz. Except that it isn't 1000Hz. It's actually some higher frequency
bizzare PC clock that is divided down to approximate 1000Hz. It has an
error of several percent since the original clock is not an integer
multiple of 1000.
I'm feeling especially charitable today, so I've attached a replacement
for ztdummy that uses the time of day clock instead. Note that your clock
should be atomic-clock synchronized with ntp for this to work perfectly,
although its performance even with a non-synched clock will be far superior
compared to the old code.
-DG
David G. Lawrence
President
Download Technologies, Inc. - http://www.downloadtech.com - (866) 399 8500
Co-Founder, The FreeBSD Project - http://www.freebsd.org
Pave the road of life with opportunities.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
Posted: Sat Jul 28, 2007 12:43 am Post subject: [Asterisk-bsd] MeetMe drift
Quote:
Attached is a slightly improved version of my ztdummy.c that should work
if system HZ is larger than 1000.
I should probably also point out that while this may help your problem,
it won't completely solve it because it is only one piece of the puzzle.
The timer inaccuracy results in extra conference latency over time, but
the other problem is that delays in scheduling the Asterisk process can
cause big and sudden jumps in latency that this new ztdummy.c won't fix.
The real problem is that Asterisk allows voice packets to back up in
its input queue, and that has an unbounded queue depth.
Attached is a patch to Asterisk's channel.c that limits the queue
depth of voice frames to a maximum of just 2 frames (40ms) on the channel
input queue. Under certain circumstances, this may result in an occasional
dropped voice frame, but that is better in my opinion than adding extra
latency to the call.
Posted: Sat Jul 28, 2007 11:38 am Post subject: [Asterisk-bsd] MeetMe drift
David,
Many thanks for the code and information you've posted regarding this!
Quote:
The real problem is that Asterisk allows voice packets to back up in
its input queue, and that has an unbounded queue depth.
This is probably what is affecting me the most. I'm running on low
power boards that are also serving up a PHP based configuration GUI.
Page loads cause 100% spikes in CPU usage and I was always surprised
that this didn't result in any degradation in audio quality. It does,
however, increase delay during conferences.
I'll try your patches out on Monday and report back.
Thanks again,
-Michael I.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
Posted: Sat Jul 28, 2007 5:46 pm Post subject: [Asterisk-bsd] MeetMe drift
On Fri, 27 Jul 2007 17:10:07 -0700, David G Lawrence wrote:
Quote:
Attached is a slightly improved version of my ztdummy.c that should
work if system HZ is larger than 1000.
david,
this is excellent. i'll check it out a little bit on the various freebsd
boxes we've got, and if it does work, i'll commit it to the zaptel-bsd
subversion repository.
first though, could someone give me a concrete serious of steps which is
able to show the meetme drift in action so we can compare the different
implementations of ztdummy ?
--
Regards, /\_/\ "All dogs go to heaven."
dinesh@alphaque.com (0 0) http://www.openmalaysiablog.com/
+==========================----oOO--(_)--OOo----==========================+
| for a in past present future; do |
| for b in clients employers associates relatives neighbours pets; do |
| echo "The opinions here in no way reflect the opinions of my $a $b." |
| done; done |
+=========================================================================+
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
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