From: Vijaykrishna (vijaykrishna@xxxxxxxxxxx)
Date: Tue Feb 08 2000 - 22:56:58 GMT-3
Hi Paul,
The scenario you explained here where the R2 is totally in area1; did you
tried pinging from R4 to any IP address in area 0 ?
The problem I see here is (don't consider the ISDN yet)......
R2 learns about area2/R4 from the backbone; since all the interarea traffic
needs to go via area0 for R4 the next hop for area2 is R1,
whereas for area0 the area2/r4 is reachable via the VL which passes through
R2 so the next hop is R2, so there is a loop between R1 and R2 !!
I have seen this practically...does any one see this problem or am I missing
something ??
thanks,
Vijay
-----Original Message-----
From: Stephens, Paul [Prof.Serv] <Paul.Andrew.Stephens@compaq.com>
To: 'ccielab@groupstudy.com' <ccielab@groupstudy.com>
Date: Tuesday, February 08, 2000 1:53 PM
Subject: FW: ospf demand circuits and virtual-links
>I think there are a couple of scenarios for this problem. The one I have
>been working on today is as follows
>
>
>Area 0 --- R1 --- Area 1 ---- R2 ------ Area 1 ----- R3 ----- Area 2 -----
R
>4
> [ [
> [---------------------ISDN----------------------[
>
>I have Virtual link from R3 to R1 across Area 1 and ip ospf demand-circuit
>of R3 only. The way I got it to work was to put both end of the ISDN in
Area
>1. It works fine that way, but if you put them area 0 the Isdn just stays
up
>all the time. If you put the ends in different area it gets really
confused.
>
>I'm going to try it now with the Isdn in another area, say Area 3 and see
>what that does. I suspect this is the one where you need multiple virtual
>links.
>
>
>Paul Stephens
>
>UK INI Network Consultant
>Networks and Systems Integration Services
>Compaq Computer Ltd
>
>*mailto:paul.andrew.stephens@compaq.com
><mailto:paul.andrew.stephens@compaq.com>
>* Mobile +44 7818 457948
>
>
>
> -----Original Message-----
> From: Curtis Phillips [mailto:phillipscurtis@netscape.net]
><mailto:[mailto:phillipscurtis@netscape.net]>
> Sent: Tuesday, February 08, 2000 17:36
> To: ccielab@groupstudy.com
><mailto:ccielab@groupstudy.com>
> Subject: ospf demand circuits and virtual-links
>
> There has been some talk about the demand-circuit command
>not repressing
> ospf hellos when configured in a virtual-link circuit. Can
>anyone verify this?
> Also, if so, under what circumstance and what is the remedy?
>
> Thanks,
>
> Curtis
>
>
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:22:53 GMT-3