From: Edwards, Andrew M (andrew.m.edwards@boeing.com)
Date: Wed Oct 12 2005 - 19:07:45 GMT-3
GS,
Here's the deal. ATM and ISDN, although not difficult, continue to
cause me to waste a lot of time that I don't need too on the exam.
Now, I did speak with Bruce Caslow the other day. His suggestion was
two-fold. He suggested I setup ppp authentication chap with encap ppp
because it makes ISDN more stable - thus reducing time looking at things
that aren't issues (never thought about that myself cause it seemed to
add complexity and thus problems to resolve). Second, he suggested I
throw this out to GS and see if others have had similar experiences to
see if there are "better" processes to hammer through ISDN/ATM.
So, I'm putting this out to GS because I want to hear your processes for
setting up ATM and ISDN (e.g. do this in this order). I just don't want
to waste anymore valuable time dealing with the "tweaky" behavior of
these L2 technologies and continue to carry the stigmata of the
"HeartBreak Candidate".
My recurring problem is this for each technology:
ISDN: ppp encapsulation.
ATM: PVC / SVC - varied. Just looking for process steps.
Here are the steps I take legacy ISDN.
R1:
Isdn switch-type basic-ni1
Dialer-list 1 protocol ip permit
Username R1 pass 0 cisco
Int bri0/0
Ip add xxxxx
Isdn spid1 <spid1> <LDN>
Isdn spid2 <spid2> <LDN>
Dialer-group 1
Dialer map ip yyyyy name R2 broadcast <dialstring>
Dialer map ip xxxxx name R2 broadcast <dialstring>
R2:
Isdn switch-type basic-ni1
Dialer-list 1 protocol ip permit
Username R1 pass 0 cisco
Int bri0/0
Ip add yyyyy
Isdn spid1 <spid1> <LDN>
Isdn spid2 <spid2> <LDN>
Dialer-group 1
Dialer map ip yyyyy name R1 broadcast <dialstring>
Dialer map ip xxxxx name R1 broadcast <dialstring>
* At this point I test connectivity with basic ping. This works
without incident as expected.
Then all heck breaks loose when I go to each interface and type "encap
ppp" and try to revalidate ICMP traffic with a new encapsulation scheme.
Here is what I see:
ICMP replies with "....." So they "time out". However, debug dialer
event shows line comes up and attaches to remote side. OSPF adjacencies
will come up and so will virtual-links. But no user traffic moves
through to validate link. If I show isdn status I'm multiframe
established on both spids. Even test call isdn int bri0/0 works... just
no user traffic.
And then onto ATM:
ATM:
I will be brief here. I see the requirements as either PVC or SVC.
PVC can either static map, use inarp, PPPoA (dialer profile or
virtual-temp)
SVC can either use CLIP or static mappings
What I'm looking for is the process used to minimize 'tweaky' behavior
of ATM. IOW, if its PVC, should I bother setting up ILMI and atm
ilmi-pvc-discovery to make sure I'm getting the PVC? That kind of
stuff. Very basic.
For either instance it comes down to this. I waste too much time
setting up these technologies because they are "tweaky" even when setup
correctly per requirements. What I'm looking for is a process people in
GS have found that takes away some of the tweaky behavior so I don't
waste any time on these.
Thanks inadvance
andy
This archive was generated by hypermail 2.1.4 : Sun Nov 06 2005 - 22:00:50 GMT-3