Hi,

I am trying to adjust the EIGRP metric of routes so a certain path is
preferred:


Router 1:

router eigrp 1
redistribute connected
network 10.0.0.0
default-metric 100000 200 255 1 1500
no auto-summary


Router 2:

router eigrp 1
network 10.0.0.0
no auto-summary


As I understand it the above config should adjust the delay metric
(from 100 to 200) and thus make router 2 the preferred path (Router 2
will have the default delay of 100).

Router 3, also running EIGRP AS 1 is seeing the both routes as equal:


Router 3:

R3#show ip eigrp topology

Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,
r - reply Status, s - sia Status

P 10.10.10.0/24, 2 successors, FD is 30720
via 192.168.1.1 (30720/28160), FastEthernet0/0
via 192.168.2.1 (30720/28160), FastEthernet1/0


Any suggestions greatly appreciated,

Paul

Your example modified the default metrics in the global EIGRP configuration.
Consider adjusting the EIGRP metrics within the interface configuration for
each interface. The commands "bandwidth xxx" and "delay xxx" help determine
the better path.
Try leaving the "router eigrp 1" section pretty plain and put the
"bandwidth" and "delay" commands within the interface configuration of each
interface of each router. Use "show ip eigrp topology x.x.x.x x.x.x.x" to
see more information on the metrics of learned routes.

-----
Scott Perry
Indianapolis, IN
-----

<thefunnel@aol.com> wrote in message
news:eae3b4b4-c885-4079-ba83-7db5fe4c79b0@z72g2000hsb.googlegroups.com...

If you change the default metrics, you must change them to the same values
for every router running that EIGRP process or your routing will get very ,
very weird. Cisco "recommends" that you do not change the default metrics.

As the previous poster said below, to change the metric on specific
interface use the "bandwidth" or "delay" on the interface. It is
recommended that the bandwidth statement is always set to the real bandwidth
of the interface because it is used to compute the load of the interface.
It is recommended that you change the delay to adjust the metric. The delay
is in 10s of ms. Setting it to a lower value will make the interface more
preferred, and setting it to a higher value will make it less preferred.
When changing it you must shut the interface and then do a "no shut" for the
value to become effective. To see the new computed metric use the command
"show ip eigrp <process> topology". If you do a search on Cisco you can
find the formula that EIGRP uses to compute the metric based upon the
bandwidth and delay of the interface.


"Scott Perry" <scott.perry@somecompany> wrote in message
news:4863f52f$0$18420$39cecf19@news.twtelecom.net. ..

I would recommend you to keep all metrics numbers the same (as it was
already described here, if you change it, it should be changed in each and
every router in your network). If you need to send router over one link,
while keep other as a backup, just use the "bandwidth" command on the
interface you would make "less preferred".

You can use the command "show ip eigrp topology <network> <mask> to see
which routes available to the destination network, and what metrics are for
each link. Then you go to one of the interfaces, and change the bandwidth to
a value less than real. For example, if you have a 1 Gigabit interface,
adding the command "bandwidth 800000" will increase metrics for the route.
Then you may clear your routing table (cle ip route *), and issue the same
command "show ip eigrp topology <network> <mask> and see how your preference
changed.

Good luck,

Mike
CCNP, CCDP, CCSP, CCVP, MCSE W2K, MCSE+I, Security+, etc.
CCIE Voice (in progress), CCIE R&S (probably next)
------
Headset Adapters for Cisco IP Phones
www.ciscoheadsetadapter.com
www.headsetadapter.com



<thefunnel@aol.com> wrote in message
news:eae3b4b4-c885-4079-ba83-7db5fe4c79b0@z72g2000hsb.googlegroups.com...