EIGRP metric formula

[aev-at-netlab]

Hi!

Тоday I analysed the full IEGRP metrics formula, which wrote in many documents:

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

Also in this docs wrote, what, if K5=0, the [K5/(reliability + K4)] term is not used. So, given the default values for K1 through K5, the composite metric calculation used by IGRP reduces to Metric = Bandwidth + Delay.

But it's not right: if K5 is near or equal zero, value of the metric aspire to or equal zero.

You are can't consider [K5/(reliability + K4)] only, if value of formula's part [K5/(reliability + K4)] is near or equal "1". Therefore, only if K5 ~(reliability + K4) ___/just as reliability ~ (K5-K4)/___ and {К4>0, K5>0, K5>=K4}, formula reduces to more simple form.

What you think about this?

[jitesh1982]

Hi

eigrp metric calculation formula is based on

bandwidth and delay , if u havent change the "K" values .

(10 to the power of 7 divide by slowest bandwidth along the path + Sum of Delays divide by 10 ) whole multiply by 256

i.e

(10000000/slowest bw + Sum of Delays/10)*256

[jsayer]

I agree that the explanation of the IGRP and EIGRP metric in all of Cisco's documentation does not work for the reasons you've pointed out. This issue pops up every now and then and I haven't ever seen a good explanation from Cisco. The only thing that I've come up with is to wonder if the last '*' in the formula should really be a '+' which would make that whole term in the formula drop out if the K5 value was 0.

In practical terms, I've never seen anyone use the full formula and you can reproduce the metric through manual calculation, so at least we know that the routers have the correct formula built into the IOS code.

Regards,

John