Jump to content

sandnaga

Members
  • Content Count

    41
  • Joined

  • Last visited

Community Reputation

3 Neutral

About sandnaga

  • Rank
    Member
  1. Amit, the scenario's that you have put up are ok. But as i suggested earlier, you shold not map the AF of Peering to whether or not they use community OR ext-community. - you need to know what are the EXT community attributes and what are the community attributes in BGP. and if you need to send these in the BGP control messages, then enable send community OR send ext.community. that would be a better way to learn this. - technically you can enable these two on all peerings, no harm done as far as your data/control plane is conserned. but its not the sane way of doing things. - Ofcourse, if you miss sending these you will land in trouble. thanks
  2. yes, single topology was pre-configured. you cannot use any topology you wish. you need to change to Multi-topology since you need ISIS to form MPLS traffic engineering database/topology in TE questions. Which is ok, because there is no way out. its all up to your discretion and judgement based on questions asked. Understanding this OR cracking this is the trickiest part in any CCIE. happy to help.. thanks.
  3. thats a trick question and unless you know proctors you cannot be sure of it. my logic: 1. its been long time. there is surely change around the corner. 2. BLR site just added 2 more seats in SP. They would want newer versions too. 3. some internal rumours of newer version already under test phase for last couple of months. so i would say very high chances of newer version in next few months (say 2 to 3 months from now) again, this is just my theory, could be 100% wrong in reality..
  4. amit, there are two parts to this question. 1. what is the objective at the end of the question. if it is that R1-R10 all routers need to ping their L1 in vrf ABC, then Yes, you need to ensure their L1 is availlable in VPNV6 of all the routers' BGP tables. 2. you shouldnt enable vpnv6/ipv6/ipv6-addresses or any other configuration for that matter, just because you are confused. this, in most cases leads to deduction of points - so ifi remember correctly, there is no AF ipv6 under vrf ABC for routers other than R1,R2,R3,R8 and R9. So you should stick to that - Without AF iov6 under vrf, there is no point in advertising L1 to BGP (you cannot) hope that helps. thanks PS: you can ping me on IM if you need more help, becasue i get alerts for IM and not the posts. that way i can respond sooner.
  5. mine was at R8.. they could potentially ask for R4,R3, R7. this scenario cannot happen in R6.. only VPNV4 labels there, so this condition doesnt apply. just static route should take care of labels at R2/R1 this is 3.1 am talking about.
  6. hi event84 BGP MDT section and thte topology WRT RR is same, no surprises. please go through 3.1 thanks PS: also understand the uniqueness in having control path and data path seperate, becasue 3.1 IAS can be done with option C as well. you need to be confident in that aspect.
  7. yes, TE is the same no surprises. but given few surprises folks are getting, i suggest you be strong there too.. there are couple of queries surrouding TE, ARabie had a topic going, please go through it. IT will help. thanks
  8. this should still have it.. [Hidden Content]
  9. ok, As i said b4 site 2 there are no surprises, you should be able to handle the eigrpv6 and set its metrics using RPL. site-1 what was complicated on the edge router, they had pre-configured MED value using a out-bound route policy. this bgp configuration was pre-existing. the question really never mentioned anything about this. it just requested for site-1 to site-2 XYZ cutomer pings. so, the trick here in BGP is, out bound route filters by default donot allow labelling bgp routes. <<<< this is what i requested ARabie to study, earlier in my post. you will get the nbrs UP, youl get the routes, but your data path will be broken. what you need to do is enable this labelling in the route policy. By saying "set mpls labels". I guess you could try to remove that MED route-policy. Am not a 100% sure if you will pass or fail. Please IM me if you need 1:1 discussion. we can setup a suitable time. thanks
  10. My complete configs for 3.1 Thats a bit tedious, 7hrs job and honestly SanjanaIE has done a great work in publishing it. to re-do that with ROI very low, i will respectfully say no. But 100%, surely going to clearly explain to you the differences i see from the 3.1 i got and the 3.1 that was published in this forum/ccie-cert. I am a bit busy this weekend, but will surely post the details ASAP. will explain MED and the complication in the preconfiguraion in my next post..
  11. V6 I have already answered, please check. V4, for site 2, no surprises. It's exactly as sanjanaie..
  12. did you redistribute the connected routes with the PE vrf? thats important, i havent seen your configs, but if you used OSPF as v6 ptotocol in CE, you need to mention redistribute OSPF include-connected. AS hemanshut mentioned, ping output is clearly saying its not reachable.
  13. also, most importantly, know your verification methods. i for one know all CLI on these boxes so was able to verify all valid paths and redundancy paths existence etc. because, a lot of my friends said they had even CSC pinging but didnt pass exam. Becuase they would have got all the path selection etc. wrong so its not just enough you aceive the final goal, but how you do it. if you have time, try shutting down redundant PE-CE links and try traceroutes, that will for sure let you know if any deplyment issues exist, assuming it can trace
  14. god forbig, but if they ask question like donot enable v6 in core or something, then GRE tunnels are your only option for ipv6.
×
×
  • Create New...