Again 5.2/5.3 and RPs

Hi all,

I realize that this question has been asked several times thus far, but once again can someone please explain :

1- Why in 5.2, the solution reaches to just 1 ACL, but at last uses 2 ACL ?!

2- Why we configure the loopback addresses with ip pim ?

3- Where is the MA ?

Cheers,

K.

 

Comments

  • Well, back from some diggings...I reckon can explain some whys...

    First off, this link offers a nice help :

    http://forum.internetworkexpert.com/ubbthreads.php/ubb/showflat/Number/15684/page/5#Post15684

    The last line of the question 5.2, makes us use "sparse-dense" mode, there is nothing about MA - we always have 2 options to use : Auto-RP or Static. Here for some reason (which is not very clear to me - could be lack of any info on MAs) - we go for the Static method.

    Also there are heaps of questions in the archive on the issues with the number of ACLs; I guess the correct answer is what the SG has come up with after one page of calculation!!! so I'd say we need just one line for 5.2 which is 225.10.0.0 0.48.255.255.

    As for the loopback..probably it's not a must for the task. Just sort of best practice...

    Your thoughts ?

     

    Cheers,

    K.

     

  • hey Kami,

     

    Glad to know that you went through that lab before i did :)

     

    i think that the SG is wrong and that the ACL overlaps toooo many addresses

     

    since you've done that a lab a while ago. did you find anything regarding that Task ?

     

    thx,

    hadi

  • Hey.  I just went through ver5 of this lab and it's starting to make better sense to me why static RPs are the only choice.  A mapping agent is bound at the prefix level, what they're asking you to do in the task is beyond the capabilities of what AutoRP announces (I haven't tried BSR, but I imagine the same applies).

    There are two issues that I can see with using the AutoRP method now:

    1) Assuming you use R3 as the MA (it's a good place for the MA based on it's position in the network), then you're doing your discovery and announcement from the same router.  Here are the results:

    *Apr  9 23:11:55.105: Auto-RP(0): Build RP-Discovery packet
    *Apr  9 23:11:55.105: Auto-RP:  Build mapping (225.8.0.0/14, RP:150.1.3.3), PIMv2 v1,

    The annoucements and RP maps that are built violate the task.

    Rack1R4#show ip pim rp map
    PIM Group-to-RP Mappings
    This system is an RP (Auto-RP)

    Group(s) 225.8.0.0/14
      RP 150.1.3.3 (?), v2v1
        Info source: 150.1.3.3 (?), elected via Auto-RP
             Uptime: 03:02:52, expires: 00:00:25

    2) The RP group mappings for the annouce filter (if you used one) will not accept the prefix version of the announcement from R4.

    R4

    ip pim send-rp-announce Loopback0 scope 16 group-list 1
    access-list 1 permit 226.37.0.0 1.8.255.255

    R3

    ip pim rp-announce-filter rp-list 4 group-list 44

    access-list 4 permit 150.1.4.4
    access-list 44 permit 226.37.0.0 1.8.255.255

    R3 sees the announcement as:

    *Apr  9 23:11:42.809: Auto-RP(0): Filtered 226.36.0.0/14 for RP 150.1.4.4

    So all Auto-RP annoucements are filtered.  As for the ACLs, the binary math in version 5 of the SG is correct, they should look like this:

    access-list 1 permit 225.10.0.0 0.48.255.25

    I think someone else mentioned Scott Morris' binary math breakdown earlier, it's a very good read for this sort of task.  My eyes naturally gravitated towards 16, but that only allows for 2 entries.  Putting it on paper makes it really easy to break down.

     

    -ryan

  • Hi all,

    I realize that this question has been asked several times thus far, but once again can someone please explain :

    1- Why in 5.2, the solution reaches to just 1 ACL, but at last uses 2 ACL ?!

    Cheers,

    K.

    Poor quality control. IE really needs an editor to QC these study guides. Somethings may just be different IOS, but basic ACL calculations... Come on!

  • We are publishing an Errata this week for all products. 

    On Aug 14, 2009, at 4:59 PM, webwidejosh wrote:

    image Kami:

    Hi all,

    I realize that this question has been asked several times thus far, but once again can someone please explain :

    1- Why in 5.2, the solution reaches to just 1 ACL, but at last uses 2 ACL ?!

    Cheers,

    K.



    Poor quality control. IE really needs an editor to QC these study guides. Somethings may just be different IOS, but basic ACL calculations... Come on!




    "Internetwork Expert - The Industry Leader in CCIE Preparation
    http://www.internetworkexpert.com

    Subscription information may be found at:
    http://www.ieoc.com/forums/ForumSubscriptions.aspx
    "

  • We are publishing an Errata this week for all products. 

    Hey... An upside if I fail my CCIE lab on Monday?  [^o)]

Sign In or Register to comment.