We have an MPLS cloud running ospf over nine sites.
the main site is configured as an ABR, Area 0. The remaining sites are configure with incrementing areas 1-8.
I just brought up Area 8. I had problems with Sprint from the beginning saying the original design showed all the sites with authentication and wtth the NSSA bit set. The only site set up this way was the last one, which I finally got straightened out and is up this morning.
All the sites have 172.20.x.x/30 serial addresses except the last site which is 172.21.x.x/30
The problem is this:
All of the remote sites show the last site in their route tables as an IA route with subnet and next hop information.
The HQ site does not show the route, I can ping the site, but I believe only because the router has the default gateway set to point into the cloud.
I can see the router in the OSPF database, but not in the route table.
Any one have any ideas?
Have you got area 8 set as NSSA? Are the others set up as normal areas?
It's a bit odd that the route is showing up in the ospf database, but not the routing table, it would do that if it could not see the next hop address.
What address can you see for the new site? Is it the /30 address or it the site's network?
Let me know and we'll get this sorted!
Please rate all posts
Thanks for the reply,
No this morning the provider had the area set as NSSA, but that has been corrected.
The address in the database shows up as the network.
The next hop comment maybe something, the serial interface is up on the new site router, but the ethernet is not right now. There is nothing connected to it, I was wondering if that might be the problem.
Is that it?
Do I understand correctly that the Ethernet at the last remote site is down? If it is down then OSPF (and other routing protocols as well) will not advertise it.
If you are saying that the serial link it up it will show up on the central router as directly connected. If the remote router has only an Ethernet and that Ethernet is down, then it is correct behavior for no routes to be in the table from the remote router.
Yes the ethernet at the remote site is down.
I knew the ethernet would not be advertised, but did not understand what is happening here:
The serial is showing up in the route table at all of the other remote sites, just not in the HQ site route table.
Attached is a partial database and route table, note the 172.21.3.136 network.
It shows up in both databases but not in the HQ site route table. It shows up in all other remnote site route tables also.
The only differnece is the HQ site is is an ABR or an ASBR, I can't remember which.
Consequently, the network is not available to the HQ site. I can ping both sides of the connecting network of the new site from the remote sites, but I cannot ping either side of the connecting network from the HQ site.
Check the following link to see if it helps you in solving the problem
Also, give us the show output of
show ip ospf database router 172.21.3.136
from both the remote and the hq. It will help in identifying the problem.
please post "show ip ospf database summary" from the HQ router and one remote site routers, as the network should be advertised as LSA3.
The routes show up on both end in th summary, I do not see where is shows then as LSA3.
But everything liiks the same on both ends as far as the show commands
I am glad that you got it figured out. It is one of the aspects (could I say "features") of OSPF that an access list/distribute list will permit or prevent routing entries from appearing in the local routing table but the entries are in the data base and are advertised to other peers.