I'm not going to argue with Chuck whether it's valid or not valid to split areas (because yes Chuck is correct, it is valid). From a network design perspective, it's bad enough for the use of a Virtual Link, but splitting the Areas is totally unnecessary.
A colleague of mine has the term "Network Masturbation" for designs like these. Basically by breaking the KISS principle (Keep It Simple Stupid) you are just begging for trouble. Just by taking a look at the network layout as it is, my suggestion was to change the Frame Relay Area 1 to a different Area (Area 3). This still allowed for the test of the Virtual Link, and constituted a better designed network. What you didn't see were my direct responses to Priscilla referring to examining the different topological databases to see exactly where these links were and whether the next logical phase of an LSA was being performed. (LSA Types 1 and 2 shows the link in Area 0, are LSA type 3's getting sent into Area 1 and 2 with the information...) -----Original Message----- From: s vermill [mailto:[EMAIL PROTECTED]] Sent: Tuesday, February 05, 2002 8:40 AM To: [EMAIL PROTECTED] Subject: Re: OSPF DR problem [7:34379] Priscilla Oppenheimer wrote: > > Remember, I think from a design point of view. I say "for some > reason > there's an Area 2" because I think it's a bad design not > because I was > surprised to see it there in the show output. ;-) Well that certainly makes sense. I thought you were surprised by the area because you were using a remote practice lab and weren't certain of the state of the entire network. Nevermind. > But thanks > for replying, > because it made me question my expectations. > > Here's what part of the network design looks like: > > ---R2---Area-1-ISDN----R8---Area-1-Ethernet > | > Area 0 | > Ethernet | > | > ---R1---Area-1-Frame Relay---R9---Area-2-Ethernet There was some back and forth about whether or not the partitioned area 1 was a problem. I think Moy says it best (RFC 2178, pgs 33 & 34)... (to save myself some typing, the discussion is centered on areas as being different colors, all meeting up with the edge of the backbone) "...When the AS topology changes, one of the areas may become partitioned. The graph of the AS will then have multiple regions of the same color (area ID). The routing in the Autonomous System will continue to function as long as these regions of the same color are connected by the single backbone region." > > When I did a "show ip route" on R9 and R8 I thought I would see > the > Ethernet LAN in Area 0. That was not a logical expectation? I > should just > see a default route on ABRs? > Unless configured as stub areas (which would preclude using them as transit areas), I would think you should see the topology of the backbone. Unfortunately, the RFC only addresses virtual links as a means to repair a partitioned backbone. It does not address providing bacbone connectivity to a non-backbone area. Nor does the RFC discuss demand circuits, which, of course, is a Cisco implementation. So there may very well be a "gottcha" in there that simply isn't addressed in the "official" OSPF documentation. I guess the answer will most likely be revealed when you revisit the remote lab and do some magic with debug and show. Regards, Scott > Thanks. > > Priscilla Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=34557&t=34379 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
