# EIGRP question.

Discussion in 'Cisco' started by Ed, Mar 16, 2007.

1. ### EdGuest

i have a question about Eigrp cost. How would R1 travel through the network
if the network had a other routes to get to R2? one route could be a T1,
another combo of router with 100M and another route T3. the Eigrp config
would be config as default without varience.

Since eirgp use Bandwidth and delay, i would assume it would route through
the T3, which is more bandwidth than the T1 and combo of 100M combo routers.
the icnd book explains the eigrp always take the lowest cost to route and
gives you the cost #. how do you come out with the cost # to figure out the
best cost path so packet can travel from R1 to R2.

ed

Ed, Mar 16, 2007

2. ### Martin GallagherGuest

On Thu, 15 Mar 2007 23:51:19 -0400, Ed wrote:

> i have a question about Eigrp cost. How would R1 travel through the network
> if the network had a other routes to get to R2? one route could be a T1,
> another combo of router with 100M and another route T3. the Eigrp config
> would be config as default without varience.
>
> Since eirgp use Bandwidth and delay, i would assume it would route through
> the T3, which is more bandwidth than the T1 and combo of 100M combo routers.
> the icnd book explains the eigrp always take the lowest cost to route and
> gives you the cost #. how do you come out with the cost # to figure out the
> best cost path so packet can travel from R1 to R2.
>

If by cost # you mean the EIGRP metric for the route, by default it's:

256 * ( 10**8/minimum BW + Total delay along the path)

--
Rgds,
Martin

Martin Gallagher, Mar 16, 2007

3. ### EdGuest

so by your division equation, you mean that if the BW is higher the
cost/metric would be for the route to get from R1 to R2 would be better
served if it goes through the combo 100M routers rather then the T1 route or
T3 route? Yes?

"Martin Gallagher" <> wrote in message
news...
> On Thu, 15 Mar 2007 23:51:19 -0400, Ed wrote:
>
>> i have a question about Eigrp cost. How would R1 travel through the
>> network
>> if the network had a other routes to get to R2? one route could be a T1,
>> another combo of router with 100M and another route T3. the Eigrp config
>> would be config as default without varience.
>>
>> Since eirgp use Bandwidth and delay, i would assume it would route
>> through
>> the T3, which is more bandwidth than the T1 and combo of 100M combo
>> routers.
>> the icnd book explains the eigrp always take the lowest cost to route and
>> gives you the cost #. how do you come out with the cost # to figure out
>> the
>> best cost path so packet can travel from R1 to R2.
>>

>
> If by cost # you mean the EIGRP metric for the route, by default it's:
>
> 256 * ( 10**8/minimum BW + Total delay along the path)
>
> --
> Rgds,
> Martin
>

Ed, Mar 16, 2007
4. ### TrendkillGuest

On Mar 16, 8:41 am, "Ed" <> wrote:
> so by your division equation, you mean that if the BW is higher the
> cost/metric would be for the route to get from R1 to R2 would be better
> served if it goes through the combo 100M routers rather then the T1 route or
> T3 route? Yes?
>
> "Martin Gallagher" <> wrote in message
>
> news...
>
> > On Thu, 15 Mar 2007 23:51:19 -0400, Ed wrote:

>
> >> i have a question about Eigrp cost. How would R1 travel through the
> >> network
> >> if the network had a other routes to get to R2? one route could be a T1,
> >> another combo of router with 100M and another route T3. the Eigrp config
> >> would be config as default without varience.

>
> >> Since eirgp use Bandwidth and delay, i would assume it would route
> >> through
> >> the T3, which is more bandwidth than the T1 and combo of 100M combo
> >> routers.
> >> the icnd book explains the eigrp always take the lowest cost to route and
> >> gives you the cost #. how do you come out with the cost # to figure out
> >> the
> >> best cost path so packet can travel from R1 to R2.

>
> > If by cost # you mean the EIGRP metric for the route, by default it's:

>
> > 256 * ( 10**8/minimum BW + Total delay along the path)

>
> > --
> > Rgds,
> > Martin

If you are talking a single hop in all cases, yes since a 100mbit lan
connection is higher bandwidth and lower latency than a 45mbit DS3.

Trendkill, Mar 16, 2007
5. ### EdGuest

did not know that. thats for your help.

ed

"Trendkill" <> wrote in message
news:...
> On Mar 16, 8:41 am, "Ed" <> wrote:
>> so by your division equation, you mean that if the BW is higher the
>> cost/metric would be for the route to get from R1 to R2 would be better
>> served if it goes through the combo 100M routers rather then the T1 route
>> or
>> T3 route? Yes?
>>
>> "Martin Gallagher" <> wrote in message
>>
>> news...
>>
>> > On Thu, 15 Mar 2007 23:51:19 -0400, Ed wrote:

>>
>> >> i have a question about Eigrp cost. How would R1 travel through the
>> >> network
>> >> if the network had a other routes to get to R2? one route could be a
>> >> T1,
>> >> another combo of router with 100M and another route T3. the Eigrp
>> >> config
>> >> would be config as default without varience.

>>
>> >> Since eirgp use Bandwidth and delay, i would assume it would route
>> >> through
>> >> the T3, which is more bandwidth than the T1 and combo of 100M combo
>> >> routers.
>> >> the icnd book explains the eigrp always take the lowest cost to route
>> >> and
>> >> gives you the cost #. how do you come out with the cost # to figure
>> >> out
>> >> the
>> >> best cost path so packet can travel from R1 to R2.

>>
>> > If by cost # you mean the EIGRP metric for the route, by default it's:

>>
>> > 256 * ( 10**8/minimum BW + Total delay along the path)

>>
>> > --
>> > Rgds,
>> > Martin

>
> If you are talking a single hop in all cases, yes since a 100mbit lan
> connection is higher bandwidth and lower latency than a 45mbit DS3.
>

Ed, Mar 16, 2007