From: Edwards, Andrew M (andrew.m.edwards@boeing.com)
Date: Wed Sep 21 2005 - 19:32:26 GMT-3
John,
The problem is this:
If the area 0 link between R3 and R4 were to drop, there would be
discontiguous area 0 links. These are the virtual links between R3-R1
and R4-R2.
Thus, Sundar is trying to resolve the discontiguous nature of this
potential failure with a third virtual link.
Andy
-----Original Message-----
From: John Matus [mailto:John.Matus@tokiom.com]
Sent: Wednesday, September 21, 2005 3:19 PM
To: sundar.palaniappan@verizon.com
Cc: cciein2006@yahoo.com; ccielab@groupstudy.com; nobody@groupstudy.com;
Mitchell, TJ
Subject: RE: Strange Routing choice with OSPF Virtual-Link
sundar,
i follow you with R1 and R2 becoming ABR's because of the virtual-links
but i'm unclear why they need to be connected via the virtual-link....is
this because they are "semi-discontiguous" <a condition i just made up
:)
>......i would think that R1 and R2 would be able to talk to each other
end-to-end through the 2 virtual-links without having to be directly
connected in area 0
Regards,
John D. Matus
Technical Support / PAS
Fujitsu Consulting
626-568-7716
John.Matus@tokiom.com
sundar.palaniappa
Sent by:
To
nobody@groupstudy "Mitchell, TJ"
.com <tmitchell@allianttech.com>
09/21/2005 02:00 ccielab@groupstudy.com,
Subject
RE: Strange Routing choice with
Please respond to OSPF Virtual-Link
sundar.palaniappa
R2 is in Area0. Do a show ip ospf border-routers on any another router
and will tell you R2 is an ABR. Virtual link between R1 & R3 and R2 & R4
makes R1 & R2 to become ABRs.
Virtual link between the two ABRs - R1 & R2 - via Area 1 is all needed
to patch the backbone connection btw R1 & R2.
Just lab it up and see the results.
HTH,
Sundar Palaniappan
CCIE #14532
"Mitchell, TJ"
<tmitchell@allian To: Sundar X.
Palaniappan/EMPL/PA/Verizon@VZNotes,
ttech.com> cciein2006@yahoo.com
Sent by: cc:
ccielab@groupstudy.com, nobody@groupstudy.com
nobody@groupstudy Subject: RE: Strange
Routing choice with OSPF Virtual-Link
.com
09/21/2005 04:23
PM
Please respond to
"Mitchell, TJ"
R2 isn't in area 0, so the idea won't work.
Unless of course you configure a tunnel to area 0 through area 2 without
using virtual-links then virtual-link R1 to R2 though the tunnel using
area 1. Then that would put the tunnel interface of R2 in area 0. But
then R2 is in area 1 as well so no virtual links would be required, but
then your going to run into the issue of learning the tunnel interfaces
through the tunnel and your tunnel will work intermittently due to
learning the destination tunnel address through the tunnel rather than
through the interface.
Everyone follow that?? :)
T.J. Mitchell
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
sundar.palaniappan@verizon.com
Sent: Wednesday, September 21, 2005 3:37 PM
To: cciein2006@yahoo.com
Cc: ccielab@groupstudy.com; nobody@groupstudy.com
Subject: Re: Strange Routing choice with OSPF Virtual-Link
Try creating a virtual link between R1 and R2 via Area 1
HTH,
Sundar Palaniappan
CCIE #14532
cciein2006@yahoo.
com To:
ccielab@groupstudy.com
Sent by: cc:
nobody@groupstudy Subject: Re: Strange
Routing choice with OSPF Virtual-Link
.com
09/21/2005 12:27
PM
Please respond to
cciein2006
Sorry the diagram didn't line up right.
Hope this is a little clearer:
Area 1
(R1)-(R2)
|AREA2|<----141.108.10.0/30
(R3)-(R4)
Area 0
This archive was generated by hypermail 2.1.4 : Sun Oct 02 2005 - 14:40:16 GMT-3