Multicasting Seminar

I have seen the multicast vseminar and its wonderful but i have a couple of questions though..

1- In sparse mode the request to join source based tree is generated by the RP itself or the pim neigbour that currently reciving the multicast...

2- The pim neigbour will join the source tree if RP is not in the path to the source...right
   My question here is how the PIM neigbour can decide whether path
   from the RP is the shortest or the path directly to the source ?

3- for a pim neigbour the path to the RP is ok, but the path to the source is having a RPF failure.

    now my question is if i will give "ip pim spt threshold infinity" , then pim neigbour    will always use the path from the RP does this avoids the RPF failure
   check against that particular source for which actually the RPF is failing?

I will recommend all of the guys to have a look on this seminar, it's worth watching..

Keep up the good work guys

Thanks

Comments

  • 1.  Originally the S,G is from the RP to the source.  Because in the beginning, the join-requestor has no idea where the source is.  But as the flow starts coming in, it can make a decision whether or not it has a better path (based on similar RPF-type check).

    2.  It MAY decide to join.  Offset by command like you list below.  Based on same mechanism as RPF check.  With *,G, the RPF is against the RP's address.  But it still SEES the actual source as the packets come in, so it makes a comparison.

    3.  With the threshold to inifinity, it will never join the S,G, therefore the *,G's RPF check will be to the RP's adddress not the actual source address.  So if you've gotten ANY packets, it should be good.  :)

    HTH,

    Scott







    ----- Original Message -----
    From: "Bored" <[email protected]>
    Sent: Tue, January 27, 2009 8:15
    Subject: [CCIE R&S] Multicasting Seminar


    I have seen the multicast vseminar and its wonderful but i have a couple of questions though..

    1- In sparse mode the request to join source based tree is generated by the RP itself or the pim neigbour that currently reciving the multicast...

    2- The pim neigbour will join the source tree if RP is not in the path to the source...right
       My question here is how the PIM neigbour can decide whether path
       from the RP is the shortest or the path directly to the source ?

    3- for a pim neigbour the path to the RP is ok, but the path to the source is having a RPF failure.

        now my question is if i will give "ip pim spt threshold infinity" , then pim neigbour    will always use the path from the RP does this avoids the RPF failure
       check against that particular source for which actually the RPF is failing?

    I will recommend all of the guys to have a look on this seminar, it's worth watching..

    Keep up the good work guys

    Thanks



    --
    View this message online at: http://ieoc.com/forums/p/4947/16170.aspx#16170
    --
    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
    <!>

Sign In or Register to comment.