Re: EIGRP unequal metric LB

Thank you very much.

This is exactly what I was looking for.

From:
"Mohamed El Henawy" <m.henawy_at_link.net>
To:
<michael.botha_at_gm.com>
Date:
2009/07/30 11:07 AM
Subject:
Re: EIGRP unequal metric LB

Check this one

http://www.ccietalk.com/2008/09/21/eigrp-unequal-cost-load-balancing

----- Original Message -----
From: <michael.botha_at_gm.com>
To: <ccie_at_reid.it>
Cc: <ccielab_at_groupstudy.com>
Sent: Thursday, July 30, 2009 11:52 AM
Subject: Re: EIGRP unequal metric LB

> Hi,
>
> Thanks for the response.
> I am trying to figure out the the maths required to achieve the last
> painful option :(
>
> Here is the output of what I am working with
>
> Rack1R4(config-if)#do sh ip route 222.22.2.0
> Routing entry for 222.22.2.0/24
> Known via "eigrp 1024", distance 170, metric 156928, type external
> Redistributing via eigrp 1024
> Last update from 174.1.145.1 on Serial0/0/0, 00:00:25 ago
> Routing Descriptor Blocks:
> 174.1.145.5, from 174.1.145.5, 00:00:25 ago, via Serial0/0/0
> Route metric is 567552, traffic share count is 11
> Total delay is 21170 microseconds, minimum bandwidth is 100000 Kbit
> Reliability 255/255, minimum MTU 1500 bytes
> Loading 1/255, Hops 1
> 174.1.145.1, from 174.1.145.1, 00:00:25 ago, via Serial0/0/0
> Route metric is 567808, traffic share count is 11
> Total delay is 21180 microseconds, minimum bandwidth is 100000 Kbit
> Reliability 255/255, minimum MTU 1500 bytes
> Loading 1/255, Hops 1
> * 174.1.45.5, from 174.1.45.5, 00:00:25 ago, via FastEthernet0/1
> Route metric is 156928, traffic share count is 40
> Total delay is 5130 microseconds, minimum bandwidth is 100000 Kbit
> Reliability 255/255, minimum MTU 1500 bytes
> Loading 1/255, Hops 1
>
> Rack1R4(config-if)#do sh ip ei topo 222.22.2.0
> IP-EIGRP (AS 1024): Topology entry for 222.22.2.0/24
> State is Passive, Query origin flag is 1, 1 Successor(s), FD is 152064
> Routing Descriptor Blocks:
> 174.1.45.5 (FastEthernet0/1), from 174.1.45.5, Send flag is 0x0
> Composite metric is (156928/25600), Route is External
> Vector metric:
> Minimum bandwidth is 100000 Kbit
> Total delay is 5130 microseconds
> Reliability is 255/255
> Load is 1/255
> Minimum MTU is 1500
> Hop count is 1
> External data:
> Originating router is 150.1.5.5
> AS number of route is 0
> External protocol is RIP, external metric is 7
> Administrator tag is 0 (0x00000000)
> 174.1.145.5 (Serial0/0/0), from 174.1.145.5, Send flag is 0x0
> Composite metric is (567552/25600), Route is External
> Vector metric:
> Minimum bandwidth is 100000 Kbit
> Total delay is 21170 microseconds
> Reliability is 255/255
> Load is 1/255
> Minimum MTU is 1500
> Hop count is 1
> External data:
> Originating router is 150.1.5.5
> AS number of route is 0
> External protocol is RIP, external metric is 7
> Administrator tag is 0 (0x00000000)
> 174.1.145.1 (Serial0/0/0), from 174.1.145.1, Send flag is 0x0
> Composite metric is (567808/25856), Route is External
> Vector metric:
> Minimum bandwidth is 100000 Kbit
> Total delay is 21180 microseconds
> Reliability is 255/255
> Load is 1/255
> Minimum MTU is 1500
> Hop count is 1
> External data:
> Originating router is 150.1.1.1
> AS number of route is 1
> External protocol is OSPF, external metric is 20
> Administrator tag is 0 (0x00000000)
>
>
>
>
>
> From:
> Andy Reid <ccie_at_reid.it>
> To:
> michael.botha_at_gm.com
> Cc:
> ccielab_at_groupstudy.com
> Date:
> 2009/07/30 10:27 AM
> Subject:
> Re: EIGRP unequal metric LB
>
>
>
> Hi Michael,
>
> I am going to need more information.
> First, what is the network and router from which you are attempting to
> perform unequal cost load balancing.
> Then, on that router, what is the output of "show ip eigrp AS topology
> IP/MASK"
> Is variance ruled out, if not, then I would be using the variance X
> command (as long as the advertised distance of the worst route is lower
> than the feasible distance).
> If it isn't an even multiplier (which it is here - x 4) then you can
> change the maximum paths also.
>
> Note that if you are not allowed to use the variance command then you
> can use an offset list to manipulate and add to the metric value (from
> the inbound interface) - though you cant use it to subtract from the
> metric value.
>
> The last option would be to manipulate bandwidth and delay values - this
> is more painful as it is slow and hard to get the right balance.
>
> regards Andy
>
> michael.botha_at_gm.com wrote:
>> Hi,
>>
>> I am really battling to calculate the delay required to get unequal
cost
> LB
>> ratio's correct.
>>
>> I have two routes to a destination prefix.
>>
>> One via a frame-relay link and the other via an ethernet segment.
>>
>> I need to balance the traffic with a ration of 4:1.
>>
>> By setting the delay I manage to get a ration of 40:11 which is pretty
> close to
>> 4:1 but not exactly. I then up the delay on one interface and turn it
> down on
>> the other and I get a ratio of 38:10.
>>
>> Anyways - bottom line is. I am missing the maths here to get these
ratio
> 100%
>> spot on.
>>
>>
>> Blogs and organic groups at http://www.ccie.net
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Thu Jul 30 2009 - 12:02:22 ART