6bone diagram changes
Bob Fink LBNL
RLFink@lbl.gov
Thu, 20 Feb 1997 18:09:26 -0800
6bone diagram version 54:
added NYU/US, USOT-ECS/UK, BT-LABS/UK, INFN-CNAF/IT
Welcome to the new sites, and my apologies to older sites I have been slow
to add to the diagram.
I am still inaccurately representing the IFB/UK homed leaf sites, as there
are actually some transits in use (I think). If BT-LABS and ENCOMIX could
hook directly to IFB it might be cleaner...then again it might not.
The IFB cluster seems to be all UK based (but for INR/RU) and it would be
nice if the UK folk got together among themselves and decided a v6 topology
that matches their v4 topology and then identified what is really backbone,
transit and leaf.
6bone backbone links diagram version 5:
added NWNET/US as a backbone site
added various static and RIPng links
We are making good progress in moving aaway from static routes. The goal
should still be for an all dynamically routed backbone with appropriate
links that match v4 topology.
One problem emerging with all the new RIPng tunnels is that lots of routes
are coming from places you might not want them. There are little or no
filters in use and the backbone is a flat one hop backbone so hop metrics
don't help you. I would encourage backbone sites to huddle on this and
carefully craft filters and other metrics/policies as appropriate to keep
their route distribution reasonable.
Thanks,
Bob