Hi Abdus,
Route tagging when redistributing into BGP is not supported. That’s why you get that error:
R1(config)#route-map SET_TAG permit 10
R1(config-route-map)#set tag 10
R1(config)#router bgp 1
R1(config-router)#redistribute ospf 1 route-map SET_TAG
% "SET_TAG" used as redistribute ospf into bgp route-map, set tag not supported
You can use something else in BGP to “tag” redistributed routes, like a community value. Between OSPF and EIGRP, you can use tags to prevent routing loops.
Let’s think about the things that could go wrong in a topology like this if you redistribute between OSPF-EIGRP, OSPF-BGP, and EIGRP-BGP:
* Everything you redistribute into OSPF will always show up as an external route. Even if a route that originated in OSPF gets redistributed in OSPF, this won’t cause any issues since an O (inter-area) route is always preferred over an external (O E1 or O E2) route. You do have to watch out if you have O E1 or O E2 routes that originated in your OSPF domain.
* Everything you redistribute into EIGRP will always show up as an external route. Even if the route originated in EIGRP, it won’t be a problem since EIGRP external routes have a higher AD than internal routes. You do have to watch out when you redistribute EIGRP external routes that originated in your EIGRP domain.
* Routes that originated in BGP could be redistributed into OSPF > EIGRP and get back to BGP, or the other way around, into EIGRP > OSPF > BGP.
If you don’t have OSPF external routes that originated in OSPF or EIGRP external routes that originated in EIGRP, then you don’t really have to worry about OSPF or EIGRP. You should ensure that BGP routes don’t loop around though.
If you redistribute from BGP into OSPF on R3/R4 then you can set a tag. Configure R5 to deny everything that is tagged when you redistribute into BGP. This prevents the EIGRP domain from learning BGP routes that originated in the BGP domain.
Do the same thing on R7, tag the routes you receive and that are redistributed into EIGRP. Configure R6 to deny routes that are tagged from being redistributed from EIGRP into BGP.
This prevents your BGP domain from receiving any prefixes that originated there. The only downside to this solution is that the EIGRP domain can’t use the OSPF as a backup path to get to the BGP domain (or the other way around, from OSPF > EIGRP > BGP).
If you want this, then you could do something like this:
- Configure R1/R2 to advertise a community value for all internal BGP prefixes to R3/R4.
- Configure R3/R4 to redistribute OSPF and set a tag for the prefixes with the community value.
- R5 can redistribute from OSPF to BGP and set a community value based on the tag.
- R6 can redistribute from BGP into EIGRP, set a tag based on the community value.
- R7 can redistribute from EIGRP into BGP, denying everything that has the tag.
This prevents from feeding BGP prefixes going from OSPF > EIGRP > BGP while allowing the EIGRP domain to use the OSPF domain to get to the BGP domain. You can do the same thing for BGP prefixes that go from R8 > R7 > R6 > R5 > R3/R4 > R1/R2.
I hope this helps! Redistribution with three redistribution points and three routing protocols can get complex 