asterisk users - Feb 2006 - dual TE410, both span 3 is broken

This afternoon I finally figured out more with regarding to a strange
clock-slip problem we have on our asterisk box.

We have two TE410s, in E1 mode:
 TE410P version c01a009b

They have their own interrupts:
 66:  781648298  783747388   IO-APIC-level  t4xxp
233:  253890977 1311504670   IO-APIC-level  t4xxp

They have their full 31 channels:
 span=3,0,0,ccs,hdb3,crc4
 bchan=63-77,79-93
 dchan=78

 span=7,0,0,ccs,hdb3,crc4
 bchan=187-201,203-217
 dchan=202

And properly configured in zapata:
 group = 3
 context=unused
 switchtype=national
 pridialplan=unknown
 prilocaldialplan=unknown
 signalling = pri_net
 channel => 63-77,79-93

 group = 7
 context=unused
 switchtype=national
 pridialplan=unknown
 prilocaldialplan=unknown
 signalling = pri_cpe
 overlapdial=no
 channel => 187-201,203-217

And on them, both the span 3 interfaces gives huge amount of HDLC
errors in the logging and clockslip-ticks on the line. Juggling the
cables between other interfaces, and the then-non-span-3 is clear
and the new-span-3 is full of HDLC errors and clockslips.


Myself I'm going with a "this must be hardware problem", unless
somebody else has an idea.


Edwin
-- 
Edwin Groothuis      |            Personal website: http://www.mavetju.org
edwin@mavetju.org    |          Weblog: http://weblog.barnet.com.au/edwin/

I've seen a similar problem before. Span 3 was throwing errors for  
(what seemed to be) no reason at all. After some testing it seemed  
that the number of errors thrown on Span 3 had a relationship to the  
temperature inside the servers. After installing additional cooling  
the errors had reduced significantly IIRC. This was across numerous  
servers each having 2 Digium quad T1/E1 cards in them.

May not be the same story in your case but try cooling the server and  
see if that helps some.

On Feb 12, 2006, at 3:04 AM, Edwin Groothuis wrote:
> This afternoon I finally figured out more with regarding to a strange
> clock-slip problem we have on our asterisk box.
>
> We have two TE410s, in E1 mode:
>  TE410P version c01a009b
>
> They have their own interrupts:
>  66:  781648298  783747388   IO-APIC-level  t4xxp
> 233:  253890977 1311504670   IO-APIC-level  t4xxp
>
> They have their full 31 channels:
>  span=3,0,0,ccs,hdb3,crc4
>  bchan=63-77,79-93
>  dchan=78
>
>  span=7,0,0,ccs,hdb3,crc4
>  bchan=187-201,203-217
>  dchan=202
>
> And properly configured in zapata:
>  group = 3
>  context=unused
>  switchtype=national
>  pridialplan=unknown
>  prilocaldialplan=unknown
>  signalling = pri_net
>  channel => 63-77,79-93
>
>  group = 7
>  context=unused
>  switchtype=national
>  pridialplan=unknown
>  prilocaldialplan=unknown
>  signalling = pri_cpe
>  overlapdial=no
>  channel => 187-201,203-217
>
> And on them, both the span 3 interfaces gives huge amount of HDLC
> errors in the logging and clockslip-ticks on the line. Juggling the
> cables between other interfaces, and the then-non-span-3 is clear
> and the new-span-3 is full of HDLC errors and clockslips.
>
>
> Myself I'm going with a "this must be hardware problem",
unless
> somebody else has an idea.
>
>
> Edwin
> -- 
> Edwin Groothuis      |            Personal website: http:// 
> www.mavetju.org
> edwin@mavetju.org    |          Weblog: http://weblog.barnet.com.au/ 
> edwin/
> _______________________________________________
> --Bandwidth and Colocation provided by Easynews.com --
>
> Asterisk-Users mailing list
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-users

On Sun, Feb 12, 2006 at 04:30:50PM -0700,
asterisk-users-request@lists.digium.com wrote:
> I've seen a similar problem before. Span 3 was throwing errors for  
> (what seemed to be) no reason at all. After some testing it seemed  
> that the number of errors thrown on Span 3 had a relationship to the  
> temperature inside the servers. After installing additional cooling  
> the errors had reduced significantly IIRC. This was across numerous  
> servers each having 2 Digium quad T1/E1 cards in them.
It got even funnier... After some careful log-file-reading, it
seemed this started to happen after the upgrade to 1.2.x. Before
it we couldn't use the the 1.0.8 and 1.0.9 version of the zaptel
drivers because it couldn't get an E1 link towards a C2821 and a
Vegastream 400. With the 1.2.x version it at least had a link, but
all these errors on the span 3...

That were "old" quad PRI cards (first generation). We ordered some
"new' quad PRI cards (2nd generation) and there it all works fine
with... Really really really makes me wonder about what went wrong
with our setup.

Edwin
-- 
Edwin Groothuis      |            Personal website: http://www.mavetju.org
edwin@mavetju.org    |          Weblog: http://weblog.barnet.com.au/edwin/