2

After reading the article about synchronization issues in BGP, I realised there's one thing I don't get.

http://www.cisco.com/c/dam/en/us/support/docs/ip/border-gateway-protocol-bgp/26634-bgp-toc16.gif

RTC in AS300 sends updates about 170.10.0.0. RTA and RTB run iBGP, so RTB gets the update and is able to reach 170.10.0.0 via next hop 2.2.2.1. Remember that the next hop is carried via iBGP. In order to reach the next hop, RTB must send the traffic to RTE. Assume that RTA has not redistributed network 170.10.0.0 into IGP. At this point, RTE has no idea that 170.10.0.0 even exists. If RTB starts to advertise to AS400 that RTB can reach 170.10.0.0, traffic that comes from RTD to RTB with destination 170.10.0.0 flows in and drops at RTE.

Isn't it the key reason of why people use BGP - the fact that IGP protocols such as OSFP don't need to know about thousands of different networks from other ASes, because BGP routers handles it for them?

The article states that RTE, which is an IGP router, should also know about 170.10.0.0 advertised to RTA over BGP by RTC. If RTC advertised other prefixes as well, that would mean RTE (IGP router!!) would have to learn them as well so that RTB can communicate with AS300.

1
  • Did any answer help you? if so, you should accept the answer so that the question doesn't keep popping up forever, looking for an answer. Alternatively, you could provide and accept your own answer.
    – Ron Maupin
    Aug 13, 2017 at 3:06

2 Answers 2

1

You are right to see the problem with RTE. This shows the difference between theory and practice. This isn't a practical problem for several reasons:

  1. This is a case of an AS being a transit AS. That is, AS 100 provides connectivity between AS 300 and 400.
  2. Many ISPs (who provide transit) run BGP on all their routers, so RTE would also be a BGP speaker.
  3. Synchronization is turned off in many cases.
  4. The vast majority of ISPs run MPLS in their core. This is a different paradigm than using an IGP to advertise reachability within the AS.

Ultimately, this isn't much of a problem because network designers have built networks that specifically avoid this issue with IGP and BGP.

4
  • With AS400 removed, suppose RTB wants to reach 170.10.0.0. RTA would have to tell RTE about networks learned from AS300 (RTC).
    – user299869
    Jan 23, 2016 at 12:12
  • In that case, RTA would typically advertise a default route into AS 100, so that all external traffic would flow to RTA without it having to advertise every route. There would be no reason for RTB to run BGP.
    – Ron Trunk
    Jan 23, 2016 at 12:31
  • By default route you mean the network 0.0.0.0/0? Every other AS100 router would attempt to send everything to RTA in this case (if it didn't have a 'better' matching route in its routing table).
    – user299869
    Jan 23, 2016 at 18:53
  • That's correct. In the scenario you proposed, if the only way out of AS100 is through RTA, that's what you want. If there are other exit points, then you would need to distribute those routes into the IGP so a better determination can be made.
    – Ron Trunk
    Jan 23, 2016 at 18:56
0

The problem is routers in AS100 doesn't know anything about 2.2.2.0/30 network, so they can't reach 170.10.0.0 even if they have the route to this network.

I think you can solve this problem in four ways:

  1. build full mesh iBGP in AS100 network
  2. implement route reflector(s) in AS100 network and establish sessions between RR and every router in AS100
  3. use next-hop self option
  4. use MPLS inside AS100

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.