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

Re: (ASCEND) speeding up callback



On Sat, Nov 29, 1997 at 08:50:51PM +0100, Dr. Wolfgang Beneicke wrote:
> Michiel Boland <boland@diva.nl> wrote:
> 
> > Hi
> > 
> > it currently takes more than 10 seconds to set up a max-pipeline
> > connection via D channel callback. This is waaaay to long for our
> > purposes.  Any tricks to make this go faster? Can I blame our
> > telco?  Max running 5.0Ap5, Pipeline 5.0Ap13 (don't tell me to
> > upgrade :) 
> > 
> > 09:44:15 ascrouter ASCEND: slot 9 port 4, Outgoing Call, 
> > 09:44:22 ascrouter ASCEND: slot 0 port 0, line 1, channel 4, Incoming Call, 
> > 09:44:22 ascrouter ASCEND: slot 9 port 3, Assigned to port, 
> > 09:44:24 ascrouter ASCEND: slot 9 port 3, line 1, channel 4, Call Connected, 
> > 09:44:24 ascrouter ASCEND: call 127 AN slot 9 port 3 64K 
> > 09:44:25 ascrouter ASCEND: call 126 CL 0K   c=2 p=2
> > 09:44:25 ascrouter ASCEND: slot 0 port 0, line 1, channel 3, No Connection, Cause code 016
> > 09:44:25 ascrouter ASCEND: slot 9 port 4, Call Terminated
> > 09:44:27 ascrouter ASCEND: slot 9 port 3, LAN session up,

10s is good ! We have cases where the outgoing call of the initial caller
is delayed to clear for 40 or 50 seconds, so getting a clear callback in
10 s is not bad. Some Q.931 readings seem to settle my impression that
7 s is the absolute minimum one should give the callback, else you will
always get a collision. Mileage may vary by switch vendor and software
version (I started to understand those telcos and their software version
problems since our 5.0Ap hopping, though...).

> That's what I see here all the time. Making connections between a
> MAX1800 and Pipelines, using CHAP. If I watch the MAX's status window I
> see
> - Connection up
> - LAN session down
> <call back>
> - LAN session up

This is not D-Channel callback then, it is Ascend proprietary callback
protocol.

> Sure I tried CallerID=Prefer. No change, i.e. it authenticates via CHAP.
> With CallerID=Require it won't authenticate at all. All numbers are
> correct, though.

Oops. IF you have the correct calling # in the profile or in RADIUS it
_will_ authenticate by CLID. Be sure to use the number as the Max sees it
incoming: with NET3/5 f.i. you get the number without the leading zero
of the prefix (if you dial 089 123456 you get 89123456 or even 891234560 if
this is an "Anlagenanschluß").

> Other users in Germany have complained earlier that D-channel signaling
> does not work here using a MAX (-kann das jemand bestaetigen?-)

Nee ;)

It is problematic. I've seen general problems with CLID auth if you have
no RADIUS auth at all (simple fix: enable Max as RADIUS server and point
it back to itself). There are general problems if a P50 or such dials
in and _wants_ to be authed by CHAP while the Max already authed it
by CLID. There are general problems with asymmetric auth (Max CHAPs P50
but P50 CLIDs Max or such) - always disappear if the same auth is used
in both directions. There even exist TRs for these problems. We have been
able to fix these problems with trial-and-error sessions of some length
up to now and have working ISDN callback with certain customers.

We currently have NO fix for the RADIUS problem we discovered in 5.0Ap33
that causes units which dial in to be refused (they worked up to p27).
These refusals are EXTREME EVIL - we have one customer who had a P50
calling for THREE DAYS at a 20 SECOND INTERVAL getting a B channel (CASH-
BELL-NOISE) and instantly beeing dropped by the Max due to this CHANGE.
"Deutsch Telecom" (cited from Ascend release notes) gained enormous bucks
on this single incident. We are back at p13 and refuse any further upgrade
until clarification of this GEBUEHREN-GAU.

ANYONE: Seen this with p36 ? It seems to happen with units trying to
establish MPP or MP while the RADIUS profile states PPP and the units
are not allowed to negotiate MP neither is MP negotiated away as it
worked until p33.

> I wouldn't be surprised if it was the same in the Netherlands.

Again, I would be glad if I could manage 10 s flawless ISDN callback
without hacks and in all environments ;)

-- 

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>


References: