Task 2.8 - Why next hop self?

Why do we set next-hop-self to R2? Without it if we drop the link R1-R3, R1 has a BGP route via 23.3 and has the 23 network via OSPF area 51 from R2? So it still works?

Comments

  • Actually it's because of the route recursion of BGP. Router 2 will receive 2 routes, one from R1 and one from R3 for VLAN5. It will choose the route from R1 because it has a better metric: 

    image

    R2 will have to recurse to 154.12.23.3. R2 use the route from OSPF.

    image

    Hence, R2 will go directly to R3 in order to get to VLAN5. One thing to remember i that BGP does not change the next when advertising routes from eBGP to iBGP.

    image

    With the next-hop-self on R1 for R2, R2 will see the next-hop for VLAN 5 as 192.10.12.1 

    image

    Then the traffic will go through R2, R3,R5

    image

    Pay attention to details and always traceroute to the destination to make sure that the traffic is flow as it is suppose to; both in the configuration and troubleshooting section.


    HTH

     

  • I'm still little bit confused. Doesn't R3 still get preferred over R1 because prefix to VLAN5 is learned from R3 as eBGP (metric 20) in contrary to R1, where it is learned as iBGP (200). So ospf path-cost to next-hop will not be involved in this ibgp/ebgp scnario. Or am I wrong? 

  • Hi Bodik,

    in this case, the next hop in BGP route is not directly connected, thus IGP -OSPF in this case- is involved to reach the BGP next-hop.

  • Pay attention to details and always traceroute to the destination to make sure that the traffic is flow as it is suppose to; both in the configuration and troubleshooting section.

    Thanks, that's really great explanation and advice with traceroute and double-checking details is also very good one.

     

  • Yep Indeed ! 

     

    There is another scenarios like these on the VOL2 LAB 1 . 

     

     

    :)

  • I had a different solution but still achieved the same result as you and the SG.....

     

    I could obviuosly see that the RIB adverticed the prefix correctly to the mentioned NH...but to get to the NH it followed over the serial instead if the Frame link....hence I've created a virtual link between R1 and R2 that routed packets via the frame-relay connection and once trace was checked it flowed via the relevant serial to the NH as question intended!

     


    Rack7R2#sh ip ospf int brie

    Interface    PID   Area            IP Address/Mask    Cost  State Nbrs F/C

    VL0          1     0               192.10.7.2/24      64    P2P   1/1

    Se0/1        1     0               154.7.23.2/24      64    P2P   1/1

    BV1          1     51              192.10.7.2/24      64    DROTH 2/3

    Lo0          1     51              150.7.2.2/24       1     LOOP  0/0


    Tracing the route to 154.7.5.5


      1 192.10.7.1 40 msec 40 msec 52 msec

      2 154.7.13.3 52 msec 56 msec 52 msec

      3 154.7.0.4 60 msec 60 msec 64 msec

      4 154.7.45.5 80 msec *  81 msec



    before the VL this was the trace output...



    Rack7R2#trace 154.7.5.5


    Type escape sequence to abort.

    Tracing the route to 154.7.5.5


      1 154.7.23.3 16 msec 16 msec 16 msec

      2 154.7.0.4 45 msec 48 msec 44 msec

      3 154.7.45.5 44 msec *  40 msec










  • Would creating a virtual link not violate the rules? (Maybe it doesn't, I don't know)

  • Would creating a virtual link not violate the rules? (Maybe it doesn't, I don't know)


    To solve this problem in the exam, if it is possible, I would not change my IGP, considering that you have verified your IGP before doing BGP task. I think using next-hop self is the simplest solution.

  • Hi,

       Ths solution with virtual link is too complicated and 100% not related to the task which is BGP. I would say you will not get the points in this case.

    Good luck with your studies!

  • I did fail in the task because I didn't have the "next-hop self" it did this mistake irrespective of reading nice blog posted by B.Dennis ( http://blog.ine.com/2011/10/20/bgp-route-preference-the-mostly-commonly-missed-task-in-rs-ccie-mock-lab-2/ ) on how important it is for us to rely on tracert verification for a BGP route rather than just looking at the BGP table and think that things will work.  I hope I won't this mistake again.

  • The task says configure As 300 yet we apply additional configuration to as 200 (next-hop-self). Isn't this a violation of the task?

  • I'm still little bit confused. Doesn't R3 still get preferred over R1 because prefix to VLAN5 is learned from R3 as eBGP (metric 20) in contrary to R1, where it is learned as iBGP (200). So ospf path-cost to next-hop will not be involved in this ibgp/ebgp scnario. Or am I wrong? 

    This was something you missed from the OSPF section. Remember the top part of the section where it stated the customer is running OSPF Area 0 everywhere, and some changes are necessary? That's a big clue to remove the network 0.0.0.0 255.255.255.255 area 0 statements on all of the OSPF routers, and nowhere in the tasks does it say to advertise VLAN5 into OSPF, nor is ti included in the area 3457 diagram. 

     

    Pull OSPF off R5 VLAN5, and then that problem resolves itself. 

     

    I do think that the task is a bit misleading, as it explicitly states to configure AS 300, not 200. 

  • I agree it does say configure AS300 but without adding the needed changes within AS200 and AS400 to bring the route into AS300 in the first place (VLAN5, although I didn't do this as it was already in from the initial configs) and fixing the routing throught the rest of the ASs we have control over it wouldn't work.

    So, AS300 is where the configuration is applied and ASs 200 and 400 are where the candidate needs to apply further controls to allow AS300's correct changes to complete a whole solution.

Sign In or Register to comment.