From: Stanford M. Wong (stanford@xxxxxxxxxxxxxx)
Date: Thu Aug 30 2001 - 02:37:57 GMT-3
here are some things that you can add to your list and others that are
interested -
Backup - you can either wait until the link fails but also when a threshold
of traffic is exceeded on your primary link. Be careful when backing up a
frame interface link though. Since back up will only monitor the physical
status of the interface, if the other side of the FR link goes down the
backup may never come up since the physical state of the line is still "up".
In this case you would have to use a subinterface with the frame-relay
interface-dlci command. that way when the other side of the frame relay
link goes down, that DLCI will go inactive thus activating the backup
interface (this is what I learned at my local cisco academy by the way).
Keep in mind that the Frame Relay DLCI does take awhile to go inactive, then
and only then with the backup interface timer kick in. I would suspect this
type of issue would also exist for ATM too, but I haven't tested it.
For distance vector routing protocols like (RIP v1, IPX RIP, IGRP) you can
use the snapshot routing. Be careful on using it on a physical or dialer
interface. It is a little different. You should practice to make it work
both ways.
For ospf you can use the ospf demand network keyword to keep the link down.
There is also the watch list which works with the watch group. When a route
is loss, the watch group can bring up the isdn link. This works with OSPF
and EIGRP too.
The good old fashion way to bring up a back up link is the good old static
routes with a distance greater. I have seen people use this with the
default and a route that they want to monitor. Of course you should know
your distances for all the routing protocols so you can make this static
route higher.
Hope this helps you...and good luck with your lab attempt...don't forget to
treat this as a positive experience....it was for me...three whole times...
stanford
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Matt Wagner
Sent: Wednesday, August 29, 2001 7:08 PM
To: ccielab@groupstudy.com
Subject: ISDN (Lab next week)
Hey folks,
I have my lab next week, and I am nervous about ISDN backup. I can
configure ISDN and troubleshoot it confidently, but in studying the docs and
the archives to find out the various ways to do DDR, I have managed to
confuse myself. What I am trying to clarify for myself (and anyone else who
needs it) is what the various methods of implementing a backup solution are
and what the basic componenets (not configuration details) of each are.
Here is what I am pretty sure of, and if anyone adds to it or corrects me I
will be grateful.
1. Backup interface. Simple enough-- the interface comes up as soon as the
primary link goes down and then the routers exchange routes. Not instant
failover, but easy to configure.
2. No backup interface/Dialer list config. This would require configuring
a dialer list that denies the routing protocol updates and permits other
sruff (IP, IPX, etc.). Obviously the primary link must be preferred over
the ISDN link. Should this then NOT flap? The problem with this is that
the local router won't know about remote routes via the ISDN link, so I dial
will never occur. I guess that I would then need snapshot routing, but if
so, what is the purpose of the dialer list denying the routing protocol in
the first place?
3. Snapshot routing. See #2 above.
I am making assumptions, of course, like "the ISDN link can't come up (flap)
until the primary link fails", etc. I know that this sounds kind of basic,
and I apologize for it being so, but although I can configure the various
pieces, I'm not clear on what constitutes a legitimate solution. It all
seems to overlap when I read up on it, and now I'm frazzled with a week to
go.
If it comes down to it, I don't mind giving up a couple of ISDN points; it
hurt me pretty badly timewise on my first attempt. But if anyone has a
concise summary or explanation to tie it all together, that would rock.
BTW, congrats on the recent CCIEs gt 8000. Your stories are inspiring, and
I hope to add my story to the archives Monday after next. We'll see...
Matt
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:32:00 GMT-3