In order for the second-best path to be used for unequal-cost load balancing using variance, it MUST meet the feasibility condition. IOW, it has to qualify to be the feasible successor route. If it doesn't, it won't matter how great you make the variance value.
Francis, look at the output of "show ip eigrp topology all" on R1. Check to see whether the "advertised distance" (after the slash) for the alternate route via R3 is LESS THAN the "feasible distance" (total metric) for this route via R2 (this one should be your successor route, the one that's best). I haven't worked out your numbers, but I have a feeling it won't meet the feasibility condition, and that's why you're not getting the unequal-cost load balancing you seek. Pamela At 12:02 AM 10/21/01 -0400, you wrote: >In this scenario, I don't know why you use 128 for variance value. In >short, you choose a value of variance such that this value multiplies by the >best path should be GREATER than the alternate path you like to load >balancing. > >For this scenario, the bandwidth on R1-R2-R4 path is 3000K, bandwidth on >R1-R3-R4 path is 1544K. Using a variance of 2 should load balance between >the 2 paths. > >1544k x 2 > 3000k > >Thomas N. > > >""kwock99"" wrote in message >[EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > > I have tried some basic testing on the EIGRP load balancing. For >simplicity, > > I only set up 4 router in order to get equal cost load balancing. > > > > > > LANA******R1 ----------R2 ----------R4******** LANB > > | | > > --------------R3------------ > > > > ***: Ethernet > > ----: Wan > > > > If I use the default setting for the bandwidth and delay, I can get two > > route from LAN A to LANB. > > > > R1--R2--R4 > > R1--R3--R4 > > > > After I changed all the bandwidth of serial interface of R1, R2, R4 to > > 3000Kbit (default is 1544kbit), I cannot get two route to LAN B, only the > > best route appears (R1--R2--R4). It is normal. > > > > I key in the "variance 128" command in R1 in order to get two route, but >it > > failed. The parameter 128 is make sure that the R1 will take any alternate > > route to LAN B because the metric of R1--R3--R4 must be less than 128 * > > (metric of R1--R2--R4). > > > > Anyone have the idea? Thanks. > > > > Francis Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=23706&t=23623 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
