Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: (ASCEND) ISDN dial out problem/q.931 tracing

On Thu, Jan 29, 1998 at 12:18:22AM -1000, Brian Uechi wrote:

First of all, I've looked it up in Q.931 and the cause code table
references to the B-channel-selection (5.1.2) where the statement
is relatively clear: 44 means the B-channel exclusively requested
was not available or, when non-exclusively requested (i.e. preferred)
_no_ channel at all is available. Sounds pretty clear ;)

> I used a combination of Andre Beck's (beck@ibh-dd.de) very useful
> dtrace program and my own manual annotations based on the q.931 doc
> to produce the following. I've also sent this to GTE for their
> analysis. I ignored all the layer 2 stuff.
> 
> PRI-XMIT-2/2: 25 octets @ B05B00E4      ;layer 3
>   [0000]: 08 02 04 5C 05 04 02 88 90 18 04 E1 80 83 17 70
>     08          <proto DSS1>
>     02 04 5C    <callref caller 045c>
>     05          <msgtype SETUP>
>     04          <Bearer Capability,
>     02          length 2,
>     88          ccitt, cap Unrestricted digital,
>     90          mode circuit, rate 64 kbps>
>     18          <Channel ident,
>     04          length 4,
>     E1          int identified, PRI/other, preferred, not D chan, chan follows

Ok, should mean "give me the channel which follows or, if unavailable, any
one you can hand out".

>     80          int id 00,

That's the interesting point. You explicitely name an interface. Are you
sure that this is the correct value here ? Q.931 makes no statements about
how long this may be and what the content really means, i.e. leaves it
to the carrier. IIRC this is called NFAS and is used to drive multiple
E1/T1 PRIs with one D, but maybe the first interface is actually 01 or
0005 or whatever. Your telco should be able to tell you ;)

>     83          ccitt, chan follows, B channel units,
>     17          channel 0x17 = 23>
>     70          <Called party number,
>   [0010]: 08 A1 3x 3x 3x 3x 3x 3x 3x
> 
>     08          length 8,
>     A1          type National, plan ISDN (E.164),
>     3x 3x 3x 3x 3x 3x 3x        # xxxxxxx

Pretty normal, except for the fact the channel ident IE has an explicit
interface reference.

> PRI-RCV-2: 13 octets @ B05AFCA0         ;layer 2 + layer 3
>   [0000]: 02 01 F8 56 08 02 84 5C 5A 08 02 82 AC
>     08          <proto DSS1>
>     02 84 5C    <callref callee 045c>
>     5A          <msgtype RELEASE COMPLETE>
>     08          <Cause,
>     02          length 2,
>     82          loc LN,
>     AC          code 2c (44)>

LN (local network, i.e. the Telco's switch on the other end of the T1)
thinks no channel at all is available. Typical case of "phone the telco,
phone the manufacturer of the telcos switch and let them fight it out"...

> The call setup looks fine to me but I haven't looked at one before we
> started having problems so I have no baseline. Any advice would be
> greatly appreciated!

As noted, besides from the explicite interface reference it looks very
normal. Either the switch thinks _all_ B's are active or it stumbles
about the interface value. I would first investigate on the latter.

-- 

Kanther-Line: PGP SSH IDEA MD5 GOST RIPE-MD160 3DES RSA FEAL32 RC4

+-o-+--------------------------------------------------------+-o-+
| o |               \\\- Brain Inside -///                   | o |
| o |                   ^^^^^^^^^^^^^^                       | o |
| o | Andre' Beck (ABPSoft) beck@ibh-dd.de XLink PoP Dresden | o |
+-o-+--------------------------------------------------------+-o-+
++ Ascend Users Mailing List ++
To unsubscribe:	send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd:	<http://www.nealis.net/ascend/faq>
Follow-Ups: References: