From: Sam Munzani (sam@xxxxxxxxxxx)
Date: Sat Sep 09 2000 - 23:18:08 GMT-3
best way to deal mutual redistribution with ospf is route maps instead of
distribution list. Distribution list brings type-5 or type-7 LSAs in
protocol.
Sam
----- Original Message -----
From: "John Conzone" <jkconzone@home.com>
To: "Aaron DuShey" <aaron.dushey@dushey-consulting.com>; "CCIE (E-mail)"
<ccielab@groupstudy.com>
Sent: Saturday, September 09, 2000 5:14 PM
Subject: Re: routing loops
> Aaron, I had a similar experience with lab 8, which I never did
rectify.
> I had my OSPF into BGP back into OSPF and yes, when I did I debug IP OSPF
> packet, my problem was an external LSA being generated everytime my isdn
> went down.
> The link would go down, would send that into BGP on 1 router, who
would
> then send that back into OSPF, who would genreate the external LSA, and
the
> cycle continued.
> If you debug ip ospf packet, you should see the route that is
generating
> the loop.
> One good thing is I am going over route redistribution in both Doyles
> and Caslows book.
> Whenever I redistribute routes mutually, I filter out the native
routes
> from the other protocol's routes. So if I redistribute rip into ospf and
> ospf back into rip, I filter out the rip routes inbound on the ospf
process,
> and the ospf routes in on the rip processes to prevent feedback. Of course
> on the lab it won't be so easy!
> Please let me know what you find out.
>
> ----- Original Message -----
> From: "Aaron DuShey" <aaron.dushey@dushey-consulting.com>
> To: "CCIE (E-mail)" <ccielab@groupstudy.com>
> Sent: Saturday, September 09, 2000 5:39 PM
> Subject: routing loops
>
>
> > as3---r6---r5----r1-----as1
> > 172.168.80.x as2 172.168.70.x
> >
> > OSPF is running only on r5
> > r1 & r6 are doing mutal redistribution and running both ospf & bgp
> > Looking at the results below my redistribution filtering isn't working
as
> > planned
> >
> > I want to make sure I understand what's happening here clearly-can
someone
> > verify please?
> >
> > r1 is redistibuting route 172.168.80.x in from BGP into OSPF, r6 ospf
then
> > adv to r5, r5ospf adv to r1 and BGP route is removed. BGP route is then
> > re-entered, and loop is formed.
> > Am I correct in saying this happens with type 5 (External) only since
they
> > are treated in a distance vector sort of manner in which they are
> advertised
> > by next-hops vs the normal LS behaviour of flooding?
> > thanks,
> >
> > 01:43:00: RT: add 172.168.80.0/24 via 137.20.60.1, bgp metric [200/0]
> > 01:43:05: RT: closer admin distance for 172.168.80.0, flushing 1 routes
> > 01:43:05: RT: add 172.168.80.0/24 via 137.20.25.2, ospf metric [110/30]
> > 01:43:10: OSPF: Rcv hello from 137.20.224.5 area 2 from Ethernet0/0
> > 137.20.25.2
> > 01:43:10: OSPF: End of hello processing
> > 01:43:11: RT: del 172.168.80.0/24 via 137.20.25.2, ospf metric [110/30]
> > 01:43:11: RT: delete subnet route to 172.168.80.0/24
> > 01:43:13: RT: add 172.168.80.0/24 via 137.20.60.1, bgp metric [200/0]
> > 01:43:16: RT: closer admin distance for 172.168.80.0, flushing 1 routes
> > 01:43:16: RT: add 172.168.80.0/24 via 137.20.25.2, ospf metric [110/30]
> > 01:43:20: OSPF: Rcv hello from 137.20.224.5 area 2 from Ethernet0/0
> > 137.20.25.2
> > 01:43:20: OSPF: End of hello processing
> > 01:43:21: RT: del 172.168.80.0/24 via 137.20.25.2, ospf metric [110/30]
> > 01:43:21: RT: delete subnet route to 172.168.80.0/24
> > 01:43:23: RT: add 172.168.80.0/24 via 137.20.60.1, bgp metric [200/0]
> > 01:43:26: RT: closer admin distance for 172.168.80.0, flushing 1 routes
> > 01:43:26: RT: add 172.168.80.0/24 via 137.20.25.2, ospf metric [110/30]
> > 01:43:30: OSPF: Rcv hello from 137.20.224.5 area 2 from Ethernet0/0
> > 137.20.25.2
> > 01:43:30: OSPF: End of hello processing
> > Aaron DuShey
> >
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:24:54 GMT-3