EIGRP Metric confusion
I'm trying to get my head around the metric calculation for EIGRP.
So the way i see it is as follows:
Default equation is:
M = [K1 * bandwidth + (K2 * bandwidth) / (256  load) + K3 * delay] * [K5 / (reliability + K4)]
ok now if we are using the default K values which are k1, k3 = 1 and k2, k4, k5 = 0 then the equation would look like this
M = [1 * bandwidth + (0 * bandwidth) / (256  load) + 1 * delay] * [0 / (reliability + 0)]
which can be simplified down to this:
M = bandwidth / (256  load) + delay
Is that correct so far?
I have been reading from this site
http://www.cisco.com/warp/public/103/eigrptoc.html#theoryofoperation
and it says that the equation can be simplified down to:
metric = bandwidth + delay
Now with my example above, the simplified equation was:
M = bandwidth / (256  load) + delay
so im assuming that on the cisco site the equation is metric = bandwidth + delay because by default EIGRP only uses bandwidth and delay to compute the metric, so this is why the (256  load) is not in the equation.... is this correct?
thanks
So the way i see it is as follows:
Default equation is:
M = [K1 * bandwidth + (K2 * bandwidth) / (256  load) + K3 * delay] * [K5 / (reliability + K4)]
ok now if we are using the default K values which are k1, k3 = 1 and k2, k4, k5 = 0 then the equation would look like this
M = [1 * bandwidth + (0 * bandwidth) / (256  load) + 1 * delay] * [0 / (reliability + 0)]
which can be simplified down to this:
M = bandwidth / (256  load) + delay
Is that correct so far?
I have been reading from this site
http://www.cisco.com/warp/public/103/eigrptoc.html#theoryofoperation
and it says that the equation can be simplified down to:
metric = bandwidth + delay
Now with my example above, the simplified equation was:
M = bandwidth / (256  load) + delay
so im assuming that on the cisco site the equation is metric = bandwidth + delay because by default EIGRP only uses bandwidth and delay to compute the metric, so this is why the (256  load) is not in the equation.... is this correct?
thanks
CCIE# 38186
showroute.net
showroute.net
Comments

rakem Member Posts: 800ah i see....
equation should be written like this
M = [K1 * bandwidth + [(K2 * bandwidth) / (256  load)] + K3 * delay] * [K5 / (reliability + K4)]
ok so now with the example from the link above, and using the default m = bandwidth + delay equation:
b=(10000000/lowest bandwidth) x 256
d=(sum of all delays) x 256
the first example they give has min bandwidth = 56 and sum of delays = 2200
so that would give us
b=(10000000/56) x 256
b=178571 x 256
b=45714176
d=2200 x 256
d=563200
M = b+d
M = 45714176 + 563200
M = 46277376
That looks right to me.... i just hate they way they write it all on one line on the cisco site, its confusing.
and once this has been done for all possible routes, the route with the lowest metric is placed in the routing table and the metric is now the feasible distance, correct?
So in the BSCI exam will there be questions where i have to actually calculate the metric?CCIE# 38186
showroute.net 
Paul Boz Member Posts: 2,620 ■■■■■■■■□□I highly doubt you'll have to calculate the metric down to the formula. The exams are more about knowing how something like bandwidth or delay values can be manipulated to influence routing than being able to recall the specific formula. The logic is that to understand the more complex question you have to know the fundamentals (the formula, for example).CCNP  CCIP  CCDP  CCNA, CCDA
CCNA Security  GSEC GCFW  GCIH  GCIA
[email protected]
http://twitter.com/paul_bosworth
Blog: http://www.infosiege.net/