01-28-2008 07:09 AM - edited 03-05-2019 08:45 PM
I've worked with the simulator and my lab all weekend trying to come up with some way to get two links of unequal value to load balance over EIGRP. I can not for the life of me get EIGRP to put both links in the IP table because the AD of the slow link is so far away from the FD of the fast link, short of making them equal values. This is very much true to real life and it bothers me that I can not make it work. I'm less concerned about testing now and more concerned about uncovering a truth.
No mater the values used for BW and Delay on either link, the difference between FD/AD for any given route is negligable. And because the Feasibility condition says AD must be less than successor's FD, that pretty much means you could never have, say, a 2:1 ratio on your links...and that means a variance of anything over 1:1 (equal cost) is impossible to achieve. That means unequal cost load balancing is nothing but a theory that can not be achieved in practical terms, am I correct? Is anyone really doing this?
How on earth does one get the AD to be less than the FD and still maintain a true ratio of the link speed?
01-28-2008 07:18 AM
I agree that i haven't seen in real life the use of variance for unequal cost loadbalancing but for testing you can always modify the metric weights to just choose the delay for metric calculation and avoid bandwidth
router eigrp
metric weights 0 0 0 1 0 0
Remember you will have to configure this on all your routers though
Though the varinace parameter works and installs the route in the RT i am very doubtful about the ratio that is used to send traffic across these links. Maybe someone can test the share ratio and confirm
Narayan
01-28-2008 07:48 AM
ah, sneaky...everything I've read in my CCNP prep for BSCI says don't play with metrics. I'm going to try what you said. How dangerous would that be in the real world? I don't want to think that some day someone will look at what I did and say, "what was she thinking???"
01-28-2008 07:57 AM
This method is best suited for labs
In the real world it is not feasible avoiding bandwidth for metric calculation :-)
Narayan
01-28-2008 08:53 PM
if you change the metrics you have to change it on all routers. if the K values don't match you won't get neighbor relationships in EIGRP.
01-29-2008 07:00 AM
Well I read in a Cisco white paper (http://www.cisco.com/warp/public/103/19.html) that the ratio is derived for each feasible link in the table by taking the highest FD to the actual FD for each link.
But let me ask something else because even in my lab this just isn't working.
Topology looks like this:
SWA >> R1
SWA >> R2
So A and B each have two paths by which to find each other. The top route (R1/R3) is 15 MB link. The bottom route (R2/R4) is 6MB.
This is the config I've placed on switch B and routers 3 and 4:
Switch B
router eigrp 64
network 10.64.0.0 0.0.255.255
no auto-summary
variance 3
Router R3
router eigrp 64
redistribute bgp 1 metric 15037 200 255 1 1500
network 10.64.0.0 0.0.255.255
no auto-summary
!
router bgp 1
no synchronization
bgp log-neighbor-changes
network 10.64.0.0 mask 255.255.0.0
network 69.138.62.244 mask 255.255.255.252
redistribute static
neighbor 69.138.62.245 remote-as 65000
neighbor 69.138.62.245 timers 12 36
no auto-summary
Router R4
router eigrp 64
redistribute bgp 6401 metric 6144 200 255 1 1500
network 10.64.0.0 0.0.255.255
no auto-summary
!
router bgp 6401
no synchronization
bgp log-neighbor-changes
network 10.64.0.0 mask 255.255.0.0
network 172.20.39.244 mask 255.255.255.252
redistribute static
neighbor 172.19.39.245 remote-as 1803
neighbor 172.19.39.245 timers 12 36
neighbor 172.19.39.245 route-map SETPATH out
no auto-summary
This config yeilds the following FD/AD values at Switch B:
P 10.1.0.0/16, 1 successors, FD is 221696, tag is 65000, serno 1431
via 10.64.1.3 (221696/221440), Vlan1
via 10.64.1.2 (467968/467712), Vlan1
Now you will notice that the bottom path has an AD so high it will NOT ever be feasible.
I've played with the values of Delay and BW and now, per your suggestion the weights. There is nothing short of EQUAL metrics that will cause FD of slow link to meet Feasibility condition.
Yesterday I played with injecting BGP into EIGPR on routers with equal metrics then changing delay/bw on the links between the router and the switch. It made absolutely no change in the calculation what so ever.
Question: I thought if I changed the bw/delay between the links eigrp would pick that up but it doesn't...how is that?
01-31-2008 06:21 AM
It's almost always best to have equal cost bandwidth links, with different delays, to make variance work. Generally speaking, this only works in the field when you have a long path, not a single link, where the minimum bandwidth ends up being the same, although the local bandwidth is different, and you can adjust the delay to make it balance the way you want. I normally only see this work when you have more than two links, as well.
HTH....
Russ
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: