EIGRP composite metric equation

Am i missing something or does the default value of K5 = 0 make the whole EIGRP composite metric equation = 0?

Comments

  • Great question - that part of the equation is only considered if K5 is NOT zero. If K5 is 0 (the default) - then this part of the equation is ignored.

  • Sorry still not clear. Even from my school days anything multiplied by zero = 0.

    Am I being too mathematical about this? Cisco's use of the phrase equation I think may be throwing me?

  • No problem...let us say it this way...

    The equation for EIGRP metric is:

    Metric = [k1 * bandwidth + (k2 *
    bandwidth)/(256 - load) + k3 * delay]

    If K5 is not equal to 0, then Cisco uses the following equation:

    Metric = [K1 * bandwidth + (K2 * bandwidth) / (256 - load) + K3 *
    delay] * [K5 / (reliability + K4)]

  • Makes better sense.

    My school teachers can rest easy!

    Thanks Anthony.

  • <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">





    By default EIGRP uses bandwidth and load to calculate its composite
    metric.  Load and reliability can also be used, or the ratio at which
    bandwidth and delay are used can be changed, by modifying the metric
    weights.  The default weighting of K1 and K3 mean that only bandwidth
    and delay are used.  Like Anthony said the the calculation is as
    follows:



    metric = [k1 * bandwidth + (k2 * bandwidth)/(256 - load) + k3 * delay]
    * [k5/(reliability + k4)]



    If k5 equals zero the second half of the equation is ignored.



    The part that the documentation doesn't make clear enough is that the
    "bandwidth" is the inverse minimum bandwidth along the path multiplied
    by 2.56 * 1012, and delay is tens of microseconds
    multiplied by 256.



    The weighting of the metrics can be seen from the show ip protocols
    command.





    Brian McGahan, CCIE #8593 (R&S/SP/Security)

    [email protected]

     

    Internetwork Expert, Inc.

    http://www.InternetworkExpert.com

    Toll Free: 877-224-8987 x 705

    Outside US: 775-826-4344 x 705

    Online Community: http://www.IEOC.com

    CCIE Blog: http://blog.internetworkexpert.com






    neza wrote:

    Makes better sense.

    My school teachers can rest easy!

    Thanks Anthony.







    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

  • I think the simplest way of working it out (if K-values are default) is in two steps :

     

    1) Change the "real" bandwidth to the inverse - which is 10,000,000/bw (kbits/s)  - so for example if its an E1 then 10,000,000/2048. Once you have this number, forget the kbs value and only use this.

    2) Do the following : metric = (bandwidth + aggregatd delay ) * 256

    ta da! You have your metric. A lot simpler than trying to work out the full equation.. Go and test it on your labs ;-)

    hth

    (edit - as brian says above - the delay is in tens of microseconds - so depending on where you get the value you may need to divide by 10!)

Sign In or Register to comment.