# Quick K Value Order Question

Member Posts: 861
Hello,
Quick question, why is it that the EIGRP metric weight K value order is different then the redistribution K value order?
metric weight = TOS Bandwidth Load Delay Reliability MTU
redistribution = Bandwidth Delay Reliability Load MTU

Thanks.
Follow my CCDE journey at www.routingnull0.com

• Mod Posts: 11,962 Mod
It really doesn't matter what order they are listed in as the equation is the same regardless.

As for why IOS is that way, I suppose you'd have to ask the developers!
An expert is a man who has made all the mistakes which can be made.
• Member Posts: 861
Yea I was referring to why the IOS is that way, it seems counter intuitive to have to memorize 2 different orders...
Follow my CCDE journey at www.routingnull0.com
• Users Awaiting Email Confirmation Posts: 322
When you modify the metric weights, you are modifying K values, so they go in order K1 -> K5

When you redistribute, you are changing metrics, not K values. Those go in order of most relevant to least relevant. It's intuitive, just not intuitively intuitive.
• Member Posts: 861
Thanks alot didnt think of it that way
Follow my CCDE journey at www.routingnull0.com
• Member Posts: 272
Dieg0M wrote: »
Hello,
Quick question, why is it that the EIGRP metric weight K value order is different then the redistribution K value order?
metric weight = TOS Bandwidth Load Delay Reliability MTU
redistribution = Bandwidth Delay Reliability Load MTU

Thanks.

To be picky, the command is:

metric weight tos k1 k2 k3 k4 k5

You are not setting bandwidth, delay, reliability or load directly, you are setting a value between 0 and 255 to plug in to the equation at the relevant bits.

However, when you redistribute or set the default redistribution metric, then you are actually setting those values.

MTU is not actually used in the calculation of the EIGRP metric. If you look at the formula which networker posted, you can see that MTU is never referenced. K5 actually effectively turns on or off the final fraction in the equation - if K5 is 0, then K4 is irrelevant, as 0 / anything = 0.
Current objective: CCNA Security
My blog: mybraindump.co.uk
• Member Posts: 586 ■■■□□□□□□□
When EIGRP was introduced into the IOS, two different engineers worked on development. And this is why the order of K values between metric weights and redistribution is a bit different in terms of order. Bob Ikid and Pasqual Derriere were the guys who did it.
I'm an Xpert at nothing apart from remembering useless information that nobody else cares about.
• Users Awaiting Email Confirmation Posts: 322
MrXpert wrote: »
When EIGRP was introduced into the IOS, two different engineers worked on development. And this is why the order of K values between metric weights and redistribution is a bit different in terms of order. Bob Ikid and Pasqual Derriere were the guys who did it.

Did you pull that one out of your Derriere?
• Member Posts: 586 ■■■□□□□□□□
powmia wrote: »
Did you pull that one out of your Derriere?

No harm in a nice little story to explain things.
I'm an Xpert at nothing apart from remembering useless information that nobody else cares about.