BGP route filtering into OSPF

[tahscolony]

I have one eBGP neighbor that I want to redistribute their routes into OSPF, and a few other eBGP neighbors I want to filter as well. The problem is the one neighbor is not summarized but mixed /24 and supernets. They are all the S3 routes from AWS.

Is there a way to have something for the neighbor inbound that can then be used for the route map that distributes into OSPF?

I have a route map

route-map BGP-OSPF permit 10
 match ip address AWS-Allowed-OSPF

the ACL had deny statements for internal networks at the top to prevent routing loops that happen when redistributing OSPF into BGP and back, followed by permits for specific networks from other neighbors. Those neighbors have only a few prefixes, the one I need help with has nearly 600 prefixes.  What I am trying to avoid is a permit any rule in the route map.

S3 also uses a JSON, so is there a better way to do this on an ASR 1001-x?

[Francesco Molino]

Hi

I'm sorry but I need more information in order to help.

The router on which you want filter is running both BGP and OSPF or the redistribution is done from another router you aren't managing?

To avaid routing loops, you can set tag when you are redistribution from OSPF to BGP and then deny these tags when you'll do invert. It's a better how to handle routing loops.

Could you please explain more in details what do you want to filter? Because I've missed maybe something. You are talking about 600 prefixes.

To filter OSPF inbound you can use distribute-list features.

Maybe a sketch of your routing design will help as well.

Thanks

[tahscolony]

Looks like there really is no way to resolve this one. Amazon has their subnets all blended together, and they do massive load balancing, so their Amazon Prime can be on several different subnets, which may also be the same subnets as S3.