From: Lachlan Kidd (lkidd@xxxxxxxxxxxxxxxxxxx)
Date: Sun May 27 2001 - 03:24:31 GMT-3
Hi Mike,
After doing a quick experiment, I have to disagree...
Without the subnets keyword, only major networks which are already in the
routing table will be redistributed. From the question, only the addresses
of the four loopback were given, none of which were 172.16.0.0/16.
Have you found otherwise ?
Regards,
Lachlan
For example...On the ASBR, redistributing from EIGRP into OSPF.
170.9.0.0/24 is subnetted, 4 subnets
D 170.9.11.0 [90/409600] via 200.200.200.9, 00:00:52,
Ethernet0<---these are the routes
D 170.9.10.0 [90/409600] via 200.200.200.9, 00:00:52, Ethernet0<---to
watch
D 170.9.9.0 [90/409600] via 200.200.200.9, 00:00:52, Ethernet0
D 170.9.12.0 [90/409600] via 200.200.200.9, 00:00:52, Ethernet0
C 200.200.200.0/24 is directly connected, Ethernet0
20.0.0.0/24 is subnetted, 1 subnets
O E2 20.20.20.0 [110/20] via 172.21.19.1, 00:00:21, Serial0
172.21.0.0/16 is variably subnetted, 7 subnets, 2 masks
O E2 172.21.19.0/24 [110/20] via 172.21.19.1, 00:00:21, Serial0
C 172.21.19.0/28 is directly connected, Serial0
O E2 172.21.18.0/24 [110/20] via 172.21.19.1, 00:00:22, Serial0
O 172.21.1.0/24 [110/65] via 172.21.19.1, 00:00:22, Serial0
On another OSPF router with the subnets keyword,
Gateway of last resort is 172.21.19.1 to network 0.0.0.0
170.9.0.0/24 is subnetted, 4 subnets
O E1 170.9.11.0 [110/69] via 172.21.19.3, 00:00:13, Serial1
O E1 170.9.10.0 [110/69] via 172.21.19.3, 00:00:13, Serial1
O E1 170.9.9.0 [110/69] via 172.21.19.3, 00:00:13, Serial1
O E1 170.9.12.0 [110/69] via 172.21.19.3, 00:00:13, Serial1
O IA 200.200.200.0/24 [110/74] via 172.21.19.1, 00:00:13, Serial1
99.0.0.0/24 is subnetted, 1 subnets
O E1 99.99.99.0 [110/79] via 172.21.19.1, 00:00:13, Serial1
20.0.0.0/24 is subnetted, 1 subnets
O E2 20.20.20.0 [110/20] via 172.21.19.1, 00:00:13, Serial1
172.21.0.0/16 is variably subnetted, 7 subnets, 2 masks
O E2 172.21.19.0/24 [110/20] via 172.21.19.1, 00:00:13, Serial1
C 172.21.19.0/28 is directly connected, Serial1
rack7r6#
Same router without the subnets keyword...
Gateway of last resort is 172.21.19.1 to network 0.0.0.0
O IA 200.200.200.0/24 [110/74] via 172.21.19.1, 00:21:50, Serial1
99.0.0.0/24 is subnetted, 1 subnets
O E1 99.99.99.0 [110/79] via 172.21.19.1, 00:18:49, Serial1
20.0.0.0/24 is subnetted, 1 subnets
O E2 20.20.20.0 [110/20] via 172.21.19.1, 00:21:50, Serial1
172.21.0.0/16 is variably subnetted, 7 subnets, 2 masks
O E2 172.21.19.0/24 [110/20] via 172.21.19.1, 00:21:50, Serial1
C 172.21.19.0/28 is directly connected, Serial1
-----Original Message-----
From: Mike Gutknecht [mailto:mike@justcisco.com]
Sent: Sunday, 27 May 2001 2:23:PM
To: 'Lachlan Kidd'; Luke M; 'Ccie (E-mail)'
Subject: RE: EIGRP to OSPF without null 0
Looking at the original requirements, I think it can be solved by leaving
off the "subnets" keyword.
router ospf 10
redistribute eigrp 1 metric 1000
If there are a bunch of subnets of 172.16.0.0/16, the only route that will
be redistributed into OSPF is the /16 classful network.
-Mike G
-----Original Message-----
From: Lachlan Kidd [mailto:lkidd@netstarnetworks.com]
Sent: Thursday, May 24, 2001 8:08 PM
To: Luke M; 'Ccie (E-mail)'
Subject: RE: EIGRP to OSPF without null 0
Well, yes and no. The purpose of the route to null0 is to prevent route
black-holing. Basically with this route in the table, if the router does not
have a more specific route, it will route the packet to a bit bucket. If you
don't have the route to null0 and you do have a default route you'll have a
lot of packets following your default route out into the wilderness.
By filtering it out, you do raise this as a problem. However, other routers
running EIGRP will still learn about the summary (same applies for OSPF) and
in that way it's functioning correctly.
I guess you can look at this in a couple of ways....real world, I'd say
definitely keep the summary to null0 in the table. Saves all sorts of grief.
In a lab environment do whatever it takes to get the situation working and
meet the requirements. Going back to your idea about not having the summary
point to null0, it can't be done automatically.
Check out Doyle Vol1 p728-729
HTH,
Lachlan
-----Original Message-----
From: Luke M [mailto:luke.mendoza@home.com]
Sent: Friday, 25 May 2001 12:21:PM
To: lkidd@netstarnetworks.com; 'Ccie (E-mail)'
Subject: RE: EIGRP to OSPF without null 0
ok if you do this aren't you defeating the whole purpose of the summary?
I think the idea is to pass the summary but not have it point to Null 0
-----Original Message-----
From: Lachlan Kidd [mailto:lkidd@netstarnetworks.com]
Sent: Thursday, May 24, 2001 5:21 PM
To: Luke M; Ccie (E-mail)
Subject: RE: EIGRP to OSPF without null 0
Hi Luke,
From my experience it's quite simple
access-l 1 deny 172.16.0.0 0.0.0.0
access-l 1 permit any
router eigrp 10
distribute-list 1 in null0
That usually works for me, let me know how you go.
Regards,
Lachlan
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Luke M
Sent: Friday, 25 May 2001 9:51:AM
To: Ccie (E-mail)
Subject: EIGRP to OSPF without null 0
Ok,
Here is a scenario that a friend told me about and I just don't get... maybe
I lack the right amount of brain cells so forgive me =)
tokyo ------- ether0 -----------------ether 0 ------------------ sydney
tokyo is running EIGRP with about 4 loopbacks 172.16.1.0, 172.16.2.0,
172.3.0, 172.4.0
Sydney is running OSPF and EIGRP. I was told to allow 172.16.0.0 ( a
summary net) into OSPF without the routing procotols generating a route to
Null 0.
Ok, so I do a summary for eigrp on interface ether 0 in Tokyo. Do a show ip
route and I see 172.16.0.0 pointing to null 0 (as it should).
How do I get rid of the null 0? I don't think this is possible.
Could someone shed some light into this? This is crazy...
Luke
**Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:30:54 GMT-3