I meant to add to this thread yesterday and didn't get around to it.
Be aware that there are bugs with PPP Multilink and "fancy" queueing (as
Cisco puts it) in IOS 12.0. Or at least when interworking 12.0 with
either 11.2 or 12.1 - I don't think I've confirmed it with 12.0 at both
ends.
By fancy queueing I think Cisco means priority queuing, custom queuing,
possibly WFQ (not sure on that), etc.
What I've seen is that with 12.0 on one end and 11.2 or 12.1 on the other
(various combinations of hardware), if PPP multilink and queueing are
configured, the PPP connection doesn't establish properly.
Of course, if you're only using a single channel, PPP multilink is usually
not required.
JMcL
----- Forwarded by Jenny Mcleod/NSO/CSDA on 09/04/2002 08:44 am -----
"Herold Heiko"
Sent by: [EMAIL PROTECTED]
08/04/2002 08:16 pm
Please respond to "Herold Heiko"
To: [EMAIL PROTECTED]
cc:
Subject: RE: Ancient Isdn bri wisdom ? [7:40586]
To everyone, thank you for your advice, what you wrote seems reasonable.
However it seems this would mean if there are only single channel
connections (64k or 56k only) [m]route-cache and fair-queue can remain
enabled. I'll try that asap.
Heiko Herold
--
-- PREVINET S.p.A. [EMAIL PROTECTED]
-- Via Ferretto, 1 ph x39-041-5907073
-- I-31021 Mogliano V.to (TV) fax x39-041-5907472
-- ITALY
> -----Original Message-----
> From: Lomker, Michael [mailto:[EMAIL PROTECTED]]
> Sent: Friday, April 05, 2002 5:31 PM
> To: [EMAIL PROTECTED]
> Subject: RE: Ancient Isdn bri wisdom ? [7:40586]
>
>
> > Now I tried to research the reason for that and really didn't
> > find any.
>
> It is done because ISDN lines are considered two physical
> paths that are
> bonded together using PPP multilink or Cisco's proprietary
> bonding. If you
> don't disable route caching then the tcp/ip conversation will
> always be
> switched out the same physical path (your big download will
> only use 1/2 of
> the 128k connection). It's easy to test that...just remove
> it sometime.
>
> no-fair queue appears to be the default for asynchronous
> interfaces on the
> later versions of IOS. To be honest, I'm not certain why
> FIFO offers better
> performance than WFQ would on asynch lines; I can't find a
> good explanation
> on CCO right now.
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=40868&t=40586
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
