# EIGRP FD value calculation--URGENT

Sep 19th, 2009

dear all,

pls find the attachment of my query, and provide your suggestions ASAP

Attachment:

Hello Sourabh,

the issue here is your desire to understand the EIGRP metric.

Here are some notes about EIGRP way to calculate metrics

256 * (sum of delays in 10s of microseconds)

256 * (500 +2000+2000) = 1152000

256 * (4500)= 1152000

bw component is the inverse of the smallest BW in the path in kbps

256 * 10^7 / 1544 = 1658031

total = 2810031

not so different from 2809856 that appears in your routers.

Again FD is the best metric from the point of view of the local node (R4)

EIGRP metric is cumulative in delay but inverse proportional to the lowest BW in path so the fact that R1's loop0 has BW 8000000..... kbps has no effect in EIGRP metric calculation.

other point is that you need to divide by 10 the value you read of delay in the various sh interface because EIGRP uses tens of microseconds

Hope to help

Giuseppe

Overall Rating: 5 (1 ratings)

## Replies

paolo bevilacqua Sat, 09/19/2009 - 05:18
• Super Gold, 25000 points or more
• Hall of Fame,

Founding Member

Please use only .txt format attachments and do not solicit for urgency - this is a free forum, if you need guaranteed support use the TAC.

Giuseppe Larosa Sat, 09/19/2009 - 12:18
• Super Silver, 17500 points or more
• Hall of Fame,

Founding Member

Hello Sourabh,

the issue here is your desire to understand the EIGRP metric.

Here are some notes about EIGRP way to calculate metrics

256 * (sum of delays in 10s of microseconds)

256 * (500 +2000+2000) = 1152000

256 * (4500)= 1152000

bw component is the inverse of the smallest BW in the path in kbps

256 * 10^7 / 1544 = 1658031

total = 2810031

not so different from 2809856 that appears in your routers.

Again FD is the best metric from the point of view of the local node (R4)

EIGRP metric is cumulative in delay but inverse proportional to the lowest BW in path so the fact that R1's loop0 has BW 8000000..... kbps has no effect in EIGRP metric calculation.

other point is that you need to divide by 10 the value you read of delay in the various sh interface because EIGRP uses tens of microseconds

Hope to help

Giuseppe