[asterisk-users] JNET's qozap, dahdi and PCI-E Quad

[Tzafrir Cohen]

On Sat, May 08, 2010 at 08:02:49PM +0200, Olivier wrote:
> 2010/5/8 Olivier <oza_4h07 at yahoo.fr>
> 
> > Hi,
> >
> > I'm trying to dial from one Asterisk box to a Patton 4638 BRI gateway.
> > I'm only getting this :
> > [May  8 15:08:24] WARNING[16797]: app_dial.c:1547 dial_exec_full: Unable to
> > create channel of type 'DAHDI' (cause 0 - Unknown)
> >
> > My setup is :
> > Asterisk 1.6.1.18
> > libpri 1.4.10.2
> > 1 Junghanns QuadBRI PCI-E express with :
> > - driver
> > http://webservices.jnetdns.de/~junghanns.net/downloads/jnet-dahdi-drivers-1.0.1.tar.gz
> > - every jumper set to TE mode
> >
> >
> > When typing "pri show spans", Asterisk replies nothing :
> > brs-dev*CLI> pri show spans
> > brs-dev*CLI> pri show version
> > libpri version: 1.4.10.2
> > brs-dev*CLI> dahdi show version
> > DAHDI Version: 2.3.0 Echo Canceller: KB1

KB1? I wonder why this is preffered to mg2. 

> >
> > Each LED is green but dahdi_scan says :
> > # dahdi_scan
> > [1]
> > active=yes
> > alarms=OK
> > description=quadBRI PCI ISDN Card 1 Span 1 [TE] (ca
> > name=ztqoz/1/1
> > manufacturer=
> > devicetype=
> > location=
> > basechan=1
> > totchans=3
> > irq=0
> > type=digital-

This means that the driver's 'type' field is not set. As are "irq",
manufacturer, devicetype and location. Those should be rather simple to
fix.

> > syncsrc=0
> > lbo=399-533 feet (DSX-1)
> > coding_opts=AMI
> > framing_opts=CCS
> > coding=AMI
> > framing=CCS
> >
> >
> > 1. Is this syncsrc=0 normal (we're in TE mode) ?
> > 2. Am I correct to hope "pri show spans" to display something ?
> > 3. Suggestions ?

Hmmm... not sure.

> >
> > Regards
> >
> > I should have said I've tried unsuccessfully with both dahdi 2.2.1.1 and
> 2.3.
> I'm really wondering if PCI-E QuadBRI are supported.

It's not supported? If so, can you provide its full vendor ID and
product ID? How does the device appear on 'lspci -vn' (that is: not only
vendor ID and procut ID, but also the subvendor ID and subproduct ID.

The newer "jnet" dirvers finally use hardhdlc rather than the faked
soft-hdlc used in previous versions. Sadly they still use the
non-standard probing method.

-- 
               Tzafrir Cohen
icq#16849755              jabber:tzafrir.cohen at xorcom.com
+972-50-7952406           mailto:tzafrir.cohen at xorcom.com
http://www.xorcom.com  iax:guest at local.xorcom.com/tzafrir