[asterisk-ss7] Poor dahdi_test results

[Shaun Ruffell]

On Wed, Sep 21, 2011 at 11:37:47AM -0700, Ryan Crowder wrote:
> I'm installing it this afternoon.  I'll report back if I notice any
> problems.
> 
> -----Original Message-----
> From: asterisk-ss7-bounces at lists.digium.com
> [mailto:asterisk-ss7-bounces at lists.digium.com] On Behalf Of Shaun
> Ruffell
> Sent: Wednesday, September 21, 2011 11:35 AM
> To: asterisk-ss7 at lists.digium.com
> Subject: Re: [asterisk-ss7] Poor dahdi_test results
> 
> On Tue, Sep 20, 2011 at 01:13:34PM -0700, Ryan Crowder wrote:
> > # chrt -f 99 dahdi_test -v -c 5
> > Opened pseudo dahdi interface, measuring accuracy...
> > 
> > 8192 samples in 8159.824 system clock sample intervals (99.607%)
> > 8192 samples in 8239.544 system clock sample intervals (100.580%)
> > 8192 samples in 8159.856 system clock sample intervals (99.608%)
> > 8192 samples in 8159.833 system clock sample intervals (99.607%)
> > 8192 samples in 8239.856 system clock sample intervals (100.584%)
> > --- Results after 5 passes ---
> > Best: 99.608 -- Worst: 99.416 -- Average: 99.531525, Difference:
> > 99.997346
> 
> Still strange. Although, since the overall result is 99.9x% range I
> do not believe you will have any problems in this configuration.
> Data is not being dropped, it is just that there is a delay when
> it's being delivered to user space or the clock on your system is
> jumping around slightly.

Ryan,

This is a heads up for you.  I was looking into something else
recently that I think explains what you were seeing in this thread.

dahdi-linux 2.5.0.1 has a regression where, for digital cards, a
master span may not be selected on initial configuration. If this is
the case, even though you have a card installed / configured, the
core timer would process your timers and they run somewhere between
a 4ms and 10ms interval depending on the HZ setting in the kernel.

http://svnview.digium.com/svn/dahdi?view=revision&revision=10205

I bet if you checkout the current 2.5 branch of DAHDI-linux and test
with that, you will see more precision in the dahdi_test results.
You could also verify this is what you were seeing by putting the
span in loopback and then taking it out to have the driver rescan
for a master span.

For what it's worth,
Shaun

-- 
Shaun Ruffell
Digium, Inc. | Linux Kernel Developer
445 Jan Davis Drive NW - Huntsville, AL 35806 - USA
Check us out at: www.digium.com & www.asterisk.org