CCIE - Internetwork Expert's Online Community

Hello Experts,

my thought process for this task was to generate the aggregations on the edge devices and then filter the prefixes inside of the AS per neighbor. Here is my config

On Rack12SR6:

config t
ip prefix-list AGGREGATE deny 129.12.0.0/16
ip prefix-list AGGREGATE deny 150.12.0.0/20
ip prefix-list AGGREGATE permit 0.0.0.0/0 le 32
rb 100
aggregate-address 129.12.0.0 255.255.0.0
aggregate-address 150.12.0.0 255.255.224.0
nei 129.12.46.4 prefix-list AGGREGATE out
end


On Rack12SW2:

config t
ip prefix-list AGGREGATE deny 129.12.0.0/16
ip prefix-list AGGREGATE deny 150.12.0.0/20
ip prefix-list AGGREGATE permit 0.0.0.0/0 le 32
rb 100
aggregate-address 129.12.0.0 255.255.0.0
aggregate-address 150.12.0.0 255.255.224.0
nei 129.12.58.5 prefix-list AGGREGATE out
end

On Rack12R2:

config t
ip prefix-list AGGREGATE deny 129.12.0.0/16
ip prefix-list AGGREGATE deny 150.12.0.0/20
ip prefix-list AGGREGATE permit 0.0.0.0/0 le 32
rb 200
aggregate-address 129.12.0.0 255.255.0.0
aggregate-address 150.12.0.0 255.255.224.0
nei 129.12.23.3 prefix-list AGGREGATE out
end

...hey guy,

..thats what i did exactly...

i guess somtimes order of operation with prefix-list out /route-map out...might hose something up...(in case you had a future requiremnt that needed route-map).....i remeber reading something that the IE crew put out not sure if it intailed prefix-list/route-map .....just be aware of order of operation is what the jest of it was...

...but I was more general with the aggregate because of time constraints (129.x.0.0/16, 150.x.0.0/16)...(and it also didnt ask for not to overlap... blah blah blah

...also, how did you generate the 150.12.0.0 for advertising?.

You are right though about the order of operation between route maps and prefix-lists.

For the loopbaks, we have 10

150.X.1.0/24

150.X.2.0/24

150.X.3.0/24

150.X.4.0/24

150.X.5.0/24

150.X.6.0/24

150.X.8.0/24

150.X.9.0/24

150.X.10.0/24

They have the first 2 octets in common. Let's break the third octet.

1=  0000 0001

2=  0000 0010

3=  0000 0011

4=  0000 0100

5=  0000 0101

6=  0000 0110

7=  0000 0111

8=  0000 1000

9=  0000 1001

10=0000 1010

As you can see, they have the first 4 bits in common hence 16+4=20. The task did not specify not to overlap any network so I guess the /16 would have been good.

danielowhite:

...also, how did you generate the 150.12.0.0 for advertising?.

I am curious about this too.  I decided that the question must only be concerned with the 129.X.0.0 routes, since my 150.X.0.0 routes have not been advertised into BGP.  I suppose I would have asked the proctor for clarification in the real lab.