Quick OSPF Challenge
Comments
-
Forsaken_GA
Member Posts: 4,024
I can't think of how to get that summary into OSPF. It's driving me nuts.
This is actually what I'm stuck on as well, but I have an idea, booting up a couple routers to test it now. If it works, it would actually solve your original question as well hehe -
Forsaken_GA
Member Posts: 4,024
well, I have a solution that gets the summary into ospf without advertising the more specifics, and which falls within the rules.
setup lo0 as 192.168.0.1 255.255.248.0
add 192.168.1.1 - 4.1 as secondary IP's to that loopback interface
That gets the /21 summary into the routing table as a connected route.
create an acl that only the summary route will match, create a route-map that matches the ip of that acl, redistribute connected subnets route-map into ospf
R1#sh ip ro
C 192.168.4.0/24 is directly connected, Loopback0
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial1/0.102
C 192.168.1.0/24 is directly connected, Loopback0
C 192.168.2.0/24 is directly connected, Loopback0
C 192.168.3.0/24 is directly connected, Loopback0
C 192.168.0.0/21 is directly connected, Loopback0
R2#sh ip ro
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial0/0.201
O E2 192.168.0.0/21 [110/20] via 10.0.0.1, 00:05:51, Serial0/0.201
At first I wasn't sure if configuring a single loopback with all the IP's was breaking Colby's rules, but he says configure loopback IP's, not loopback interfaces, so unless it's a loophole and I'm just being a semantics whore, it's legal hehe
This is, of course, an entirely unsuitable application for the real world, as in a real world scenario, the loopback networks are more likely than not going to be separate interfaces, which means configuring the /21 on a loopback wouldn't be possible, as it'd cause an overlap. -
ColbyG
Member Posts: 1,264
Forsaken_GA wrote: »well, I have a solution that gets the summary into ospf without advertising the more specifics, and which falls within the rules.
setup lo0 as 192.168.0.1 255.255.248.0
add 192.168.1.1 - 4.1 as secondary IP's to that loopback interface
That gets the /21 summary into the routing table as a connected route.
create an acl that only the summary route will match, create a route-map that matches the ip of that acl, redistribute connected subnets route-map into ospf
R1#sh ip ro
C 192.168.4.0/24 is directly connected, Loopback0
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial1/0.102
C 192.168.1.0/24 is directly connected, Loopback0
C 192.168.2.0/24 is directly connected, Loopback0
C 192.168.3.0/24 is directly connected, Loopback0
C 192.168.0.0/21 is directly connected, Loopback0
R2#sh ip ro
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial0/0.201
O E2 192.168.0.0/21 [110/20] via 10.0.0.1, 00:05:51, Serial0/0.201
At first I wasn't sure if configuring a single loopback with all the IP's was breaking Colby's rules, but he says configure loopback IP's, not loopback interfaces, so unless it's a loophole and I'm just being a semantics whore, it's legal hehe
This is, of course, an entirely unsuitable application for the real world, as in a real world scenario, the loopback networks are more likely than not going to be separate interfaces, which means configuring the /21 on a loopback wouldn't be possible, as it'd cause an overlap.
Brilliant! I was trying to figure out how to accomplish something just like this, but I couldn't quite get to it. -
Forsaken_GA
Member Posts: 4,024
yeah, the challenge is getting the /21 into the routing table so it can be injected.
Although now that I think about it, I probably could have done this easier... all he said was that we couldn't use the solution to your original lab, but static routes are allowed. So I guess generating a static route for the /21 and pointing it to a loopback interface would work as well, but frankly, that's just splitting hairs and is effectively the same thing as a static to null0 -
ColbyG
Member Posts: 1,264
Forsaken_GA wrote: »yeah, the challenge is getting the /21 into the routing table so it can be injected.
Although now that I think about it, I probably could have done this easier... all he said was that we couldn't use the solution to your original lab, but static routes are allowed. So I guess generating a static route for the /21 and pointing it to a loopback interface would work as well, but frankly, that's just splitting hairs and is effectively the same thing as a static to null0
Yea, I think that would be considered the same as the first solution. -
Forsaken_GA
Member Posts: 4,024
I can think of another way to do this as well.
Create the 4 loopback interfaces on R6, to represent real networks.
Don't advertise them into ospf.
On R1, create a static route for the summary pointing to R6, and then redistribute that into OSPF.
on R6, create a route map to filter out the summary and apply into to incoming ospf updates
R6 won't have the summary route, but it doesn't need it, R1 will have the static route, and all downstream routers will have the injected summary route. -
acidsatyr
Member Posts: 111
Good thinking Forsaken, only i said you can't use redistribution. So you can't redistribute connected
Sorry if i wasn't clear enough yeah its splitting hairs lol.
So, no static point to /21 and redistribution is allowed.
And actually the solution is pretty simple you just aren't thinking about it -
ColbyG
Member Posts: 1,264
Good thinking Forsaken, only i said you can't use redistribution. So you can't redistribute connected
Sorry if i wasn't clear enough yeah its splitting hairs lol.
So, no static point to /21 and redistribution is allowed.
And actually the solution is pretty simple you just aren't thinking about it
The solution you PMed me was very far from simple. -
acidsatyr
Member Posts: 111
But it is a simple concept, you just have to know what's going on. And i was pretty sure you'll get it once i mentioned CCIP.
-
Forsaken_GA
Member Posts: 4,024
Good thinking Forsaken, only i said you can't use redistribution. So you can't redistribute connected
Sorry if i wasn't clear enough yeah its splitting hairs lol.
So, no static point to /21 and redistribution is allowed.
And actually the solution is pretty simple you just aren't thinking about it
No, you said I couldn't use the same solution as I did before, redistributing a static to null0 is vastly different than redistributing a static route pointing to a valid IP or a connected interface hehe, that does not translate to a blanket ban on redistribution.
I'm sure I'm going to kick myself when you finally post the solution, but I think I have another way to do this. -
acidsatyr
Member Posts: 111
I think i mentioned in one of posts you can't use redistribution at all; but let's say you are right, i shouldv'e been more clear. The challange is to advertise /21 into ospf.
-
Forsaken_GA
Member Posts: 4,024
ah, so you did.
alright, i give, pop me the answer
the only way I can think to do this is to define static routes for the summary on all of the routers in the area, and that is a horrible, horrible method. Especially for something that's an ospf challenge, not a generic routing challenge. -
Forsaken_GA
Member Posts: 4,024
Actually, I take that back, I did manage to get it work.
setup lo0 as 192.168.0.1/21, all other 4 IP's as secondary addresses
router ospf 1
network 192.168.0.1 0.0.0.0 area 0
R1#sh ip ro
C 192.168.4.0/24 is directly connected, Loopback0
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial1/0.102
C 192.168.1.0/24 is directly connected, Loopback0
C 192.168.2.0/24 is directly connected, Loopback0
C 192.168.3.0/24 is directly connected, Loopback0
C 192.168.0.0/21 is directly connected, Loopback0
R2#sh ip ro
Gateway of last resort is not set
10.0.0.0/24 is subnetted, 1 subnets
C 10.0.0.0 is directly connected, Serial0/0.201
O 192.168.0.0/21 [110/65] via 10.0.0.1, 00:00:05, Serial0/0.201
R2#
R2#
R2#ping 192.168.1.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 192.168.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 56/59/61 ms
R2#ping 192.168.2.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 192.168.2.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 56/58/64 ms
R2#ping 192.168.3.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 192.168.3.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 56/58/61 ms
R2#ping 192.168.4.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 192.168.4.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 56/58/61 ms
When I tried this the first time, I used
network 192.168.0.0 0.0.7.255 area 0
and it was advertising all ip's on the loopback, i apparently had a brain fart and forgot to define it exactly and require full bits to match.
It is extremely simple, and like my previous solution, wholly in apppropriate if those 4 networks are supposed to take a 'real' interface each. -
acidsatyr
Member Posts: 111
Hehe yep there you go. This is even simpler than my solution, which i realised soon after (forgot about seconderies in OSPF to be honest).
My solution was to use vrf on R6. 4 loops go to one random vrf, while yet another loop is actual /21 network. -
Forsaken_GA
Member Posts: 4,024
Hehe yep there you go. This is even simpler than my solution, which i realised soon after (forgot about seconderies in OSPF to be honest).
My solution was to use vrf on R6. 4 loops go to one random vrf, while yet another loop is actual /21 network.
Ok, I can see that, but that's an awfully long way to go just to propagate a summary route hehe. I can't think of a single instance where that would pass Occam's razor for deployment in the real world (assuming this is a companies internal network, anyway), and then there's the logistical considerations (can the router support that? Is it worth the resources?)
Still, wacky solutions like this are good I guess, never know what the lab exam will throw at you -
acidsatyr
Member Posts: 111
No this scenario should not be found in the real world (this dimension that is). But i assume we all know scenarios like this are just a mind straching exercise.
BTW vrf solution is really simple, like i said all you have to do is remember what you need to have overlaping addresses on same router! And you reminded me of secondaries. Good stuff. -
CCIEWANNABE
Banned Posts: 465
this is what i would do after looking at this task for about 3 minutes and not looking at anyones responses.
1) configure a static route to null0 to cover all networks, (192.168.0.0/21) to act as summary.
2) redistribute the static route into ospf.
3) problem solved. -
acidsatyr
Member Posts: 111
CCIEWANNABE wrote: »this is what i would do after looking at this task for about 3 minutes and not looking at anyones responses.
1) configure a static route to null0 to cover all networks, (192.168.0.0/21) to act as summary.
2) redistribute the static route into ospf.
3) problem solved.
You can't do that. -
CCIEWANNABE
Banned Posts: 465
I assume he's talking about the first one (mine).
yeah, i am talking about the first example. just labbed it up to be sure and it does work.
acidsatyr, i should have been more specific on which example i was talking about, sorry for the misunderstanding. -
acidsatyr
Member Posts: 111
CCIEWANNABE wrote: »yeah, i am talking about the first example. just labbed it up to be sure and it does work.
acidsatyr, i should have been more specific on which example i was talking about, sorry for the misunderstanding.
ups, my bad as well.
