Re: A routing problem and possible solutions (fwd)

MegaZone (megazone@livingston.com)
Tue, 26 Nov 1996 19:11:51 -0800 (PST)

Once upon a time Marc Rouleau shaped the electrons to say...
>I think so. It'll definitely handle single-host dialup connections, but
>what will happen when a non-OSPF network is connecting to us? I trust we

I know that RIP routes can be propigated into an OSPF network with
translation.

>will be able to establish static routes on our Portmasters and redistribute
>them into OSPF?

I believe static routes are advertised too, but not 100% positive.

>> Yes. This has come up before. I believe there is an RFE open for a 'black
>> hole' route.
>I guess that means that the ComOS 3.5 OSPF implementation lacks this feature?

I'm not as familiar with OSPF as I am with RIP/ARP/etc so I'm not sure of
all the tricks it provides.

We had a request with RIP to be able to say "route these packets to /dev/null"
so you could set a host or subnet black hole route for say the entire address
pool - but with a high metric. Then any that were active would take
precedence, while addresses not in use would be funnelled to nowhereland
instead of hitting the default route. Which in many cases sets up a
loop-to-infinity.

>If so, how will we handle dialup host routes -- will we be able to aggregate
>them into /27 subnets on a per-Portmaster basis even when not all ports are
>in use? Or will they have to propagate as host routes?

My take would be to use OSPF on the network and have the individual host
routes propigate. Since there is no static route on the router, it doesn't
create a loop, the router is a dead end for unused addresses.

-MZ

--
Livingston Enterprises - Chair, Department of Interstitial Affairs
Phone: 800-458-9966 510-426-0770 FAX: 510-426-8951 megazone@livingston.com
For support requests: support@livingston.com  <http://www.livingston.com/> 
Snail mail: 4464 Willow Road, Pleasanton, CA 94588