RE: Appletalk tunnels posted 11/16/2000
Definitely a good idea.
I still have 3 1/2 weeks to go until my attempt #2.
I'm just about at the point when I will just work on my speed at setting up
IP networks/addressing etc.
I spent too much time last time sorting out the addressing scheme and
setting up OSPF/BGP...
Problem is I keep finding interesting stuff to do - like all the Firewall
features, VoIP Gateways/Gatekeepers.....I'm looking forward to when I can
concentrate on new fun stuff rather than a thinner level of a lot of things.
From: Sam Munzani [mailto:sam@xxxxxxxxxxx]
Sent: Friday, November 17, 2000 9:31 AM
To: Simon Baxter; Bert Kellerman
Cc: Brian; cciemail
Subject: Re: Appletalk tunnels
Looks like you have done enough lab practices. Don't over do it now Simon
otherwise stuff like this happens. I am stopping now till tab on saturday
morning. No more lab thinking.
> I get annoyed that I don't remember things like this. I even have hand
> written notes about AT which state an AURP net must be fully meshed.
> Similarly, when ever I configure SRB I forget to put the source-route
> spanning on the token interface!!
> too much information in my head!
> -----Original Message-----
> From: Bert Kellerman [mailto:bertke@xxxxxxxxxxx]
> Sent: Friday, November 17, 2000 7:21 PM
> To: Simon Baxter
> Cc: Brian; cciemail
> Subject: Re: Appletalk tunnels
> I recreated this problem, did some research, and this is what I found.
> The RFC on AURP(1504) states this:
> "An exterior router connected to multiple tunnels functions logically as
> though it were two or more exterior routers connected to the same
> AppleTalk network, with each exterior router connected to a different
> So this IS a split horizon issue...a router will not advertise Exterior
> Routes it learns on one tunnel out the other tunnel. It views the AURP
> tunnels as the same network.
> The solution:
> Full AURP Tunnel Mesh
> Create a tunnel on each spoke pointing to the other spoke. After
> implementing this spoke to spoke route table convergence was achieved.
> (RTP on 11/29)
> To unsubscribe from the CCIELAB list, send a message to
> majordomo@xxxxxxxxxxxxxx with the body containing:
> unsubscribe ccielab
To unsubscribe from the CCIELAB list, send a message to
majordomo@xxxxxxxxxxxxxx with the body containing: