OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: Daniel Bauer (DNBZURICH.IBM.COM)
Date: Tue Apr 03 2001 - 01:53:15 CDT

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    Hello all,

    I fail to understand how this draft supports multiaccess links.
    Consider the following scenario: routers R1, R2 and R3 are connected
    to a switched ethernet N1. Then, the following directed links exist:
    R1 -> N1, R2 -> N1, R3 -> N1 and
    N1 -> R1, N1 -> R2, N1 -> R3.
    All of those links may offer a different amount of available bandwidth.
    Assuming that initially, each link offers 100 available bandwidth.
    Assuming then that an LSP is set up from R1 to R2, consuming 10
    and another LSP from R3 to R2, consuming another 20.
    Then, the available bandwidth on each link is as follows:
    R1 -> N1 : 90
    R2 -> N1 : 100
    R3 -> N1 : 80
    N1 -> R1 : 100
    N1 -> R2 : 70
    N1 -> R3 : 100

    From the draft, I get the impression that only the availabe bandwidth
    for links R1->N1, R2->N1, R3->N1 are advertised, but not the ones from
    N1->R1, N1->R2, N1->R3. Am I wrong here? Shouldn't the designated router
    advertise TE info for the N1->Rx links? If the DR advertises N1->Rx
    info,
    how does it compute the available bandwidth? If the DR does not
    advertise
    anything, how does R1 conclude that it has 90 avBw left to R3, but only
    70 to R2?

    Thanks in advance for your answers,

    -Daniel