From: Tony Olzak (aolzak@xxxxxxxxxxxxxxxxxxx)
Date: Sat Sep 30 2000 - 11:58:32 GMT-3
I realize this. I was just demonstrating to Mark how this all worked. It's
important to know in case you're ever asked to get it to work without map
statements.
Tony
----- Original Message -----
From: "Maljure, Sanjay" <smaljure@cibernetworks.com>
To: "'Tony Olzak'" <aolzak@buckeye-express.com>; "mark salmon"
<masalmon@cisco.com>; <ccielab@groupstudy.com>
Sent: Saturday, September 30, 2000 3:22 AM
Subject: RE: loop back interfaces and ospf
> Hi Tony
> What you are seeing here is spoke-to-spoke reachability provided by the L3
> routing protocol i.e. OSPF Basically the "ip ospf network
> point-to-multipoint" creates host routes to the peer and as part of the
> normal OSPF route exchange process, you are learning the /32 route to the
> other spoke with the next-hop for this route set to your hub. That is how
> you are reaching spoke2 from spoke1. But if u removed the routing
protocol,
> u would lose connectivity to the other spoke. The way to renew this
> connectivity would be to create a frame-map. Of course this is assuming
that
> spoke1 and spoke2 are on the same subnet
> Sanjay
>
> -----Original Message-----
> From: Tony Olzak [mailto:aolzak@buckeye-express.com]
> Sent: Saturday, September 30, 2000 2:46 AM
> To: mark salmon; ccielab@groupstudy.com
> Subject: Re: loop back interfaces and ospf
>
>
> Without using map statements on the spokes, while using "ip ospf network
> point-to-multipoint" on every interface, this was one of the routing
tables
> I was able to get off a spoke router:
>
> 172.17.0.0/16 is variably subnetted, 5 subnets, 2 masks
> O 172.17.59.2/32 [110/64] via 172.17.59.2, 00:00:00, Serial0
> O 172.17.59.3/32 [110/128] via 172.17.59.2, 00:00:00, Serial0
> C 172.17.59.0/28 is directly connected, Serial0
> O 172.17.59.1/32 [110/128] via 172.17.59.2, 00:00:00, Serial0
> O 172.17.59.16/28 [110/1626] via 172.17.59.2, 00:00:00, Serial0
> dopey#
>
> Note the /32 routes that say next hop as the hub router (172.17.59.2). I
did
> not use map statements on any of the spokes yet was able to get
connectivity
> from spoke to spoke.
>
> Tony
> ----- Original Message -----
> From: "Tony Olzak" <aolzak@buckeye-express.com>
> To: "mark salmon" <masalmon@cisco.com>; <ccielab@groupstudy.com>
> Sent: Friday, September 29, 2000 9:51 PM
> Subject: Re: loop back interfaces and ospf
>
>
> > Huh. That's funny. When I set mine up using the "ip ospf network
> > point-to-multipoint" command on every interface, I get a host route that
> > points to the hub router for each spoke. I'll try it again and save the
> > configs.
> >
> >
> > Tony
> >
> > ----- Original Message -----
> > From: "mark salmon" <masalmon@cisco.com>
> > To: <ccielab@groupstudy.com>
> > Sent: Friday, September 29, 2000 7:57 PM
> > Subject: Re: loop back interfaces and ospf
> >
> >
> > > I disagree. I was unable to get spoke to spoke cimmunication until I
> > > used map statements. Inverse ARP disabled)
> > >
> > > Tony Olzak wrote:
> > > >
> > > > It's not really a drawback. In a frame-relay, hub and spoke,
> multipoint
> > > > design, the spokes don't need map statements to each other because
the
> > hub
> > > > will have a host route to each spoke. It will then flood this route
to
> > every
> > > > spoke on the back bone with the next hop being itself.
> > > >
> > > > Tony
> > > >
> > > > ----- Original Message -----
> > > > From: "mark salmon" <masalmon@cisco.com>
> > > > To: <ccielab@groupstudy.com>
> > > > Sent: Friday, September 29, 2000 5:56 PM
> > > > Subject: Re: loop back interfaces and ospf
> > > >
> > > > > It is my understanding that ip ospf network point-to-multipoint's
> > > > > drawback is it advertises the interface as a host route not a
> newwork
> > > > > route (i.e. /32 mask)
> > > > >
> > > > > Stylen wrote:
> > > > > >
> > > > > > goto interfeace config mode for the loopback interface and do a
> 'ip
> > ospf
> > > > > > network point-to-multipoint' then ospf will advertise ot with
the
> > > > correct
> > > > > > mask. I think this is available in 11.3 and higher, not sure of
> the
> > > > version
> > > > > > exactly.
> > > > > >
> > > > > > Richard Foltz, CCNP, CCNP-Voice, CCDP, MCSE+I, Network+, A+
> > > > > > Technical Solutions Consultant
> > > > > > Sprint ENS
> > > > >
> > > > > --
> > > > >
> > > > >
> > > > >
> > > > > Mark Salmon
> > > > > Network Support Engineer
> > > > > SBC OP HQ
> > > > > Cisco Systems Inc
> > > > > 8735 W. Higgins Road
> > > > > Suite 300
> > > > > Chicago IL 60631
> > > > > Phone:773-695-8235
> > > > > Pager:800-365-4578
> > > > > email: masalmon@cisco.com
> > > > > Empowering The Internet Generation.
> > > > >
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:25:10 GMT-3