> On 11/21/2011 06:59 PM, Jeff Bacon wrote: > > Is there some better way to handle this? Or do I just do the > > virtual-links/dual-connects and accept the hack? > > Do you actually need areas? How many routes are involved?
There's probably 500 routes or so; it's hard to be sure entirely because many of them hide behind summarization/range-statements. It COULD all be run as a single area 0.... but given that the entire mesh spans everything from a 10G NYC metro ring to a trans-Pac internet VPN mesh, the result would seem fairly ugly. Other than the problem of "how to avoid split-area syndrome when there are > 1 ABRs joining an area to area 0 without creating separate links between the devices for area X and 0", it works fairly well as-is. (Well, that and, "why can't you tell an ABR to stop advertising the range statement when you've lost all other neighbors in that area?" but that's a fringe case.) > Could you consider a design with OSPF/iBGP or similar? That might be an ideal end-game. I still need to finish flushing out EIGRP, though, and it's a continuously-in-flux network - we keep adding sites and kit and vendor connections seemingly as fast as we have time to string it all up. _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
