I need my fellow CCNP candidates..OSPF NSSA...
networker...as a result of your comment "unless the ASBR is the ABR also", I thought "I've never labbed that", although I had read it. I figured I'd make it happen..came up with something interesting.
Here goes. Instead of explaining the topology..I scanned my notes on this. This was just supposed to be a quick lab, and the intent was that I'd write down on a separate piece of paper, what I expect the route tables to show after being fully configured..to sort of hone my skills.
http://i265.photobucket.com/albums/ii216/Mrock4/testtopo.jpg
I don't know if you can see it clear enough, but R6 and R4 are an NSSA area (Area 1). R6 has an interface coming off of it, which is addressed as 10.1.1.1/30, and that runs to R4's interface, which is addressed as 10.1.1.2. Also, I created a loopback interface on R6, to test..which is addressed as 1.1.1.1.
Here is the routing table of R3, which is in Area 0.
1.0.0.0/32 is subnetted, 1 subnets
O IA 1.1.1.1 [110/257] via 172.16.1.6, 00:00:16, Serial2/0
172.15.0.0/32 is subnetted, 1 subnets
C 172.15.0.0 is directly connected, Loopback1
172.16.0.0/16 is variably subnetted, 4 subnets, 2 masks
O 172.16.1.8/30 [110/192] via 172.16.1.6, 00:00:19, Serial2/0
C 172.16.1.4/30 is directly connected, Serial2/0
C 172.16.1.6/32 is directly connected, Serial2/0
O 172.16.1.0/30 [110/128] via 172.16.1.6, 00:00:19, Serial2/0
10.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
O IA 10.1.1.0/24 [110/256] via 172.16.1.6, 00:00:16, Serial2/0
O E2 10.1.1.1/32 [110/500] via 172.16.1.6, 00:00:09, Serial2/0
The weird part (to me, at least) ???? I don't understand why 10.1.1.0 is listed as an Interarea route, then the actual interface is listed as an external route. I understand Type 7s are translated to Type 5's..but this isn't a route outside the AS. To further confuse me, I made that 1.1.1.1 loopback, and advertised it..and it shows up as an interarea route as well. It seems just that Serial2/0 interface on R6 is showing it as external. Why?
When I debug ospf, it shows that the route is being translated at R4 from type 7 to type 5..if this is the case, how come that loopback doesn't appear as an E2 route?
ALSO..note the metric on the E2 route for that interface..500. Why, you ask? I have R4 redistributing RIP with a metric of 500. RIP is not running on R6. It is/was running on R4.....
OK, it's one of those days I guess. After typing all the stuff above, and checking and double checking, I found my problem before submitting this post. I figured maybe someone else might learn a little lesson from my mistake. Any guesses what I did wrong?.............................
I accidently left RIP (somehow) advertising on the same interface (on R4) that OSPF was running. As a result, it was seeing 10.1.1.0 as an external route (Thus giving it a metric of 500, and rightfully advertising it as an E2 route). This also explains why the 1.1.1.1 loopback was advertised (correctly) as an IA route..it didn't fall under the RIP statement. How come the 10.1.1.2 network didn't get caught as well, and appear as E2? Because my OSPF statement specifically matched the interface address, and thus having a lower AD, chose to use that..and voila, appeared as IA.
I'm officially retarded..attention to detail is key. It's funny how a stupid little mistake can make things really screw up. Oh well, I was watching Rambo while testing this out, so...I might have been a bit distracted.
Here goes. Instead of explaining the topology..I scanned my notes on this. This was just supposed to be a quick lab, and the intent was that I'd write down on a separate piece of paper, what I expect the route tables to show after being fully configured..to sort of hone my skills.
http://i265.photobucket.com/albums/ii216/Mrock4/testtopo.jpg
I don't know if you can see it clear enough, but R6 and R4 are an NSSA area (Area 1). R6 has an interface coming off of it, which is addressed as 10.1.1.1/30, and that runs to R4's interface, which is addressed as 10.1.1.2. Also, I created a loopback interface on R6, to test..which is addressed as 1.1.1.1.
Here is the routing table of R3, which is in Area 0.
1.0.0.0/32 is subnetted, 1 subnets
O IA 1.1.1.1 [110/257] via 172.16.1.6, 00:00:16, Serial2/0
172.15.0.0/32 is subnetted, 1 subnets
C 172.15.0.0 is directly connected, Loopback1
172.16.0.0/16 is variably subnetted, 4 subnets, 2 masks
O 172.16.1.8/30 [110/192] via 172.16.1.6, 00:00:19, Serial2/0
C 172.16.1.4/30 is directly connected, Serial2/0
C 172.16.1.6/32 is directly connected, Serial2/0
O 172.16.1.0/30 [110/128] via 172.16.1.6, 00:00:19, Serial2/0
10.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
O IA 10.1.1.0/24 [110/256] via 172.16.1.6, 00:00:16, Serial2/0
O E2 10.1.1.1/32 [110/500] via 172.16.1.6, 00:00:09, Serial2/0
The weird part (to me, at least) ???? I don't understand why 10.1.1.0 is listed as an Interarea route, then the actual interface is listed as an external route. I understand Type 7s are translated to Type 5's..but this isn't a route outside the AS. To further confuse me, I made that 1.1.1.1 loopback, and advertised it..and it shows up as an interarea route as well. It seems just that Serial2/0 interface on R6 is showing it as external. Why?
When I debug ospf, it shows that the route is being translated at R4 from type 7 to type 5..if this is the case, how come that loopback doesn't appear as an E2 route?
ALSO..note the metric on the E2 route for that interface..500. Why, you ask? I have R4 redistributing RIP with a metric of 500. RIP is not running on R6. It is/was running on R4.....
OK, it's one of those days I guess. After typing all the stuff above, and checking and double checking, I found my problem before submitting this post. I figured maybe someone else might learn a little lesson from my mistake. Any guesses what I did wrong?.............................
I accidently left RIP (somehow) advertising on the same interface (on R4) that OSPF was running. As a result, it was seeing 10.1.1.0 as an external route (Thus giving it a metric of 500, and rightfully advertising it as an E2 route). This also explains why the 1.1.1.1 loopback was advertised (correctly) as an IA route..it didn't fall under the RIP statement. How come the 10.1.1.2 network didn't get caught as well, and appear as E2? Because my OSPF statement specifically matched the interface address, and thus having a lower AD, chose to use that..and voila, appeared as IA.
I'm officially retarded..attention to detail is key. It's funny how a stupid little mistake can make things really screw up. Oh well, I was watching Rambo while testing this out, so...I might have been a bit distracted.
Comments
-
networker050184
Mod Posts: 11,962 Mod
Ahhh the joys of the lab! It gets frustrating, but nothing like figuring out your mistakes. Its a lot more fulfilling then when everything works the way it should.An expert is a man who has made all the mistakes which can be made. -
cisco_trooper
Member Posts: 1,441 ■■■■□□□□□□
networker050184 wrote:Ahhh the joys of the lab! It gets frustrating, but nothing like figuring out your mistakes. Its a lot more fulfilling then when everything works the way it should.
Bahh. There's alot to be said for struggle. If it works right the first time you try it then you probably aren't doing anything you didn't already know how to do...
mrock > i figured you had already tackled OSPF from where it is in the BSCI book. If you want to hurt yourself I think I have one heck of a dynamips set up for this...
Nevermind, lost it when the laptop croaked. Anyway, Doyle rules with OSPF, and once again, I found I didn't fully understand it until I read what Doyle had to say about it.. -
Mrock4
Banned Posts: 2,359 ■■■■■■■■□□
networker050184 wrote:Ahhh the joys of the lab! It gets frustrating, but nothing like figuring out your mistakes. Its a lot more fulfilling then when everything works the way it should.
It is. If things work perfect (or, if they're screwed up, and you don't know any different), you don't learn much. When they screw up, you have to really rethink all of the concepts surrounding what you're doing.
I got some caffeine, and I'm about to hammer away at some new stuff tonight. -
networker050184
Mod Posts: 11,962 Mod
Sorry, what I meant its a lot more fullfilling when you get it to work like it should after figuring out your mistakes. If its all good from the begining you don't really learn anything.An expert is a man who has made all the mistakes which can be made.
-
Mrock4
Banned Posts: 2,359 ■■■■■■■■□□
cisco_trooper..I already have a decent hold on OSPF, but in another thread (the thread title was like: OSPF...N1 and N2 routes", something made me want to test it, so I did. I'm glad I did. Most of my real world experience is with OSPF, but compared to some topologies I've set up a lab for, they weren't anything.
I'm kind of jumping around in my studies to be honest. I've finished the BSCI book minus 2 chapters: BGP and IS-IS, but I can only postpone those for so long. I've already read some IS-IS, but haven't applied it yet. I'm getting there. My buddy just passed the BSCI, and said to really have multicasting/IPv6 down, so those are something I'm trying to go over a few times. I'm not just reading through things once and calling it a day..reading chapters twice, applying what I learn, then re-reading trouble sections..and finally doings hands on again to ensure I know my stuff.