1 2 Previous Next 15 Replies Latest reply: Jan 4, 2018 1:04 AM by Prashant RSS

    EIGRP unequal load balancing

    Ping

      I labbed EIGRP unequal load balancing on the two links between R1 and R5. It's working, but only for some subnets. From R5, it's working for networks beyond R1, but not for 10.1.x.x subnets on R1. And R1 isn't doing it for 192.168.4.0 on R5 either. Why?

       

      Also, The "00:11:02" in "D        10.1.1.0 [90/156160] via 172.16.4.1, 00:11:02, FastEthernet0/0" means eigrp has been up for 11 min and 2 secs on this link?

       

       

      EIGRP unequal load balancing.jpg

       

       

      R5#show ip route

       

      Gateway of last resort is not set


            10.0.0.0/24 is subnetted, 3 subnets

      D        10.1.1.0 [90/156160] via 172.16.4.1, 00:11:02, FastEthernet0/0

      D        10.1.2.0 [90/156160] via 172.16.4.1, 00:11:02, FastEthernet0/0

      D        10.1.3.0 [90/156160] via 172.16.4.1, 00:11:02, FastEthernet0/0

            172.16.0.0/16 is variably subnetted, 7 subnets, 2 masks

      D        172.16.1.0/24 [90/3074560] via 172.16.5.1, 00:11:02, FastEthernet1/0

                            [90/2172416] via 172.16.4.1, 00:11:02, FastEthernet0/0

      D        172.16.2.0/24 [90/3074560] via 172.16.5.1, 00:11:02, FastEthernet1/0

                            [90/2172416] via 172.16.4.1, 00:11:02, FastEthernet0/0

      D        172.16.3.0/24 [90/3074560] via 172.16.5.1, 00:11:02, FastEthernet1/0

                            [90/2172416] via 172.16.4.1, 00:11:02, FastEthernet0/0

      C        172.16.4.0/24 is directly connected, FastEthernet0/0

      L        172.16.4.2/32 is directly connected, FastEthernet0/0

      C        172.16.5.0/24 is directly connected, FastEthernet1/0

      L        172.16.5.2/32 is directly connected, FastEthernet1/0

      D    192.168.1.0/24 [90/3202560] via 172.16.5.1, 00:11:02, FastEthernet1/0

                          [90/2300416] via 172.16.4.1, 00:11:02, FastEthernet0/0

            192.168.4.0/24 is variably subnetted, 2 subnets, 2 masks

      C        192.168.4.0/24 is directly connected, Loopback0

      L        192.168.4.1/32 is directly connected, Loopback0

      R5#


      ----


      R1#show ip route

       

      Gateway of last resort is not set


            10.0.0.0/8 is variably subnetted, 6 subnets, 2 masks

      C        10.1.1.0/24 is directly connected, Loopback0

      L        10.1.1.1/32 is directly connected, Loopback0

      C        10.1.2.0/24 is directly connected, Loopback1

      L        10.1.2.1/32 is directly connected, Loopback1

      C        10.1.3.0/24 is directly connected, Loopback2

      L        10.1.3.1/32 is directly connected, Loopback2

            172.16.0.0/16 is variably subnetted, 10 subnets, 2 masks

      C        172.16.1.0/24 is directly connected, Serial3/0

      L        172.16.1.1/32 is directly connected, Serial3/0

      C        172.16.2.0/24 is directly connected, Serial3/1

      L        172.16.2.1/32 is directly connected, Serial3/1

      C        172.16.3.0/24 is directly connected, Serial3/2

      L        172.16.3.1/32 is directly connected, Serial3/2

      C        172.16.4.0/24 is directly connected, FastEthernet0/0

      L        172.16.4.1/32 is directly connected, FastEthernet0/0

      C        172.16.5.0/24 is directly connected, FastEthernet1/0

      L        172.16.5.1/32 is directly connected, FastEthernet1/0

      D    192.168.1.0/24 [90/2297856] via 172.16.1.2, 00:59:08, Serial3/0

      D    192.168.4.0/24 [90/156160] via 172.16.4.2, 00:59:08, FastEthernet0/0

      R1#

        • 1. Re: EIGRP unequal load balancing
          Ing_Percy

          Hi!

          eigrp-routing-table.jpg

          About Unequal load balancing

          "If a path is not a feasible successor, the path is not used in load balancing."

          source: How Does Unequal Cost Path Load Balancing (Variance) Work in IGRP and EIGRP? - Cisco


          Best regards!

          • 2. Re: EIGRP unequal load balancing
            Ping

            Hi Ing,

             

            I'm glad you said that. And the toplogy table does have the second path for 10.1.x.x networks. So they are feasible successors. What could be other reasons?

             

            R5#

            R5#show ip ei to

            EIGRP-IPv4 Topology Table for AS(90)/ID(192.168.4.1)

            Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,

                  r - reply Status, s - sia Status


            P 172.16.5.0/24, 1 successors, FD is 2562560

                    via Connected, FastEthernet1/0

            P 10.1.3.0/24, 1 successors, FD is 156160

                    via 172.16.4.1 (156160/128256), FastEthernet0/0

                    via 172.16.5.1 (2690560/128256), FastEthernet1/0

            P 172.16.2.0/24, 1 successors, FD is 2172416

                    via 172.16.4.1 (2172416/2169856), FastEthernet0/0

                    via 172.16.5.1 (3074560/2169856), FastEthernet1/0

            P 10.1.2.0/24, 1 successors, FD is 156160

                    via 172.16.4.1 (156160/128256), FastEthernet0/0

                    via 172.16.5.1 (2690560/128256), FastEthernet1/0

            P 192.168.1.0/24, 1 successors, FD is 2300416

                    via 172.16.4.1 (2300416/2297856), FastEthernet0/0

                    via 172.16.5.1 (3202560/2297856), FastEthernet1/0

            P 172.16.3.0/24, 1 successors, FD is 2172416

                    via 172.16.4.1 (2172416/2169856), FastEthernet0/0

                    via 172.16.5.1 (3074560/2169856), FastEthernet1/0

            P 192.168.4.0/24, 1 successors, FD is 128256

                    via Connected, Loopback0

            P 172.16.1.0/24, 1 successors, FD is 2172416

                    via 172.16.4.1 (2172416/2169856), FastEthernet0/0

                    via 172.16.5.1 (3074560/2169856), FastEthernet1/0

            P 172.16.4.0/24, 1 successors, FD is 28160

                    via Connected, FastEthernet0/0

            P 10.1.1.0/24, 1 successors, FD is 156160

                    via 172.16.4.1 (156160/128256), FastEthernet0/0

                    via 172.16.5.1 (2690560/128256), FastEthernet1/0


            R5#

            • 3. Re: EIGRP unequal load balancing
              Ing_Percy

              Hi!

               

              What is your value of variance? Configurations?

              • 4. Re: EIGRP unequal load balancing
                Ping

                R5#show run | sec eigrp

                router eigrp 90

                variance 3

                network 172.16.4.2 0.0.0.0

                network 172.16.5.2 0.0.0.0

                network 192.168.4.1 0.0.0.0

                R5#


                ---


                R1#

                R1#show run | sec eigrp

                router eigrp 90

                variance 3

                network 10.1.1.0 0.0.0.255

                network 10.1.2.0 0.0.0.255

                network 10.1.3.0 0.0.0.255

                network 172.16.1.0 0.0.0.255

                network 172.16.2.0 0.0.0.255

                network 172.16.3.0 0.0.0.255

                network 172.16.4.1 0.0.0.0

                network 172.16.5.1 0.0.0.0

                R1#

                • 5. Re: EIGRP unequal load balancing
                  Ing_Percy

                  HI!

                   

                  In this network:

                   

                  P 10.1.3.0/24, 1 successors, FD is 156160

                          via 172.16.4.1 (156160/128256), FastEthernet0/0

                          via 172.16.5.1 (2690560/128256), FastEthernet1/0

                   

                  Variance:

                  FD(successor)*3 = 156160 *3 = 468 480 < 2 690 560

                   

                  For that reason it is not included in the routing table.

                   

                  Best regards!

                  • 6. Re: EIGRP unequal load balancing
                    Ping

                    OK. I fixed it with a varience of 18. Anything below 18 wouldn't work.

                     

                    Does this number affect how much traffic going through the second path?

                     

                    Gateway of last resort is not set


                          10.0.0.0/24 is subnetted, 3 subnets

                    D        10.1.1.0 [90/2690560] via 172.16.5.1, 00:00:24, FastEthernet1/0

                                      [90/156160] via 172.16.4.1, 00:00:24, FastEthernet0/0

                    D        10.1.2.0 [90/2690560] via 172.16.5.1, 00:00:24, FastEthernet1/0

                                      [90/156160] via 172.16.4.1, 00:00:24, FastEthernet0/0

                    D        10.1.3.0 [90/2690560] via 172.16.5.1, 00:00:24, FastEthernet1/0

                                      [90/156160] via 172.16.4.1, 00:00:24, FastEthernet0/0

                    • 7. Re: EIGRP unequal load balancing
                      Ing_Percy

                      Hi!

                       

                      You must apply the command: "show ip route 10.1.1.0" and see the part of "traffic share count"

                      Here is in case of CLN that explained about it:

                      EIGRP Traffic Share Count Calculation is confusing me

                       

                      Best regards!

                      • 8. Re: EIGRP unequal load balancing
                        Ping

                        Hi Ing,

                         

                        Thanks for helping.

                         

                        Below is what I get. The traffic share count is 7 and 120. So one path gets 7/127 of the traffic and the other one gets 120/127?

                         

                        R5#show ip route 10.1.1.1

                        Routing entry for 10.1.1.0/24

                          Known via "eigrp 90", distance 90, metric 156160, type internal

                          Redistributing via eigrp 90

                          Last update from 172.16.5.1 on FastEthernet1/0, 07:54:18 ago

                          Routing Descriptor Blocks:

                            172.16.5.1, from 172.16.5.1, 07:54:18 ago, via FastEthernet1/0

                              Route metric is 2690560, traffic share count is 7

                              Total delay is 5100 microseconds, minimum bandwidth is 1000 Kbit

                              Reliability 255/255, minimum MTU 1500 bytes

                              Loading 1/255, Hops 1

                          * 172.16.4.1, from 172.16.4.1, 07:54:18 ago, via FastEthernet0/0

                              Route metric is 156160, traffic share count is 120

                              Total delay is 5100 microseconds, minimum bandwidth is 100000 Kbit

                              Reliability 255/255, minimum MTU 1500 bytes

                              Loading 1/255, Hops 1

                        R5#

                        • 9. Re: EIGRP unequal load balancing
                          Ing_Percy

                          HI!

                           

                          No.

                          The ratio of traffic over each path is 7:120. It means that for each 7 packets in the first link, it is sends 43 packets in the other link.

                          In maths, the ratio is presented by: 7k/120k, where "k" is a constant integer greater than zero.

                           

                          Best regards!

                          • 10. Re: EIGRP unequal load balancing
                            Ping

                            OK. I'll try to figure out how you arrived at these numbers.

                             

                            BTW, is this still CCNP material? If not, I don't care

                            • 11. Re: EIGRP unequal load balancing
                              Ing_Percy

                              Hi!

                               

                              The calculation of these values are made aproximatelly by maths

                              156160/2690560 = 0.058039

                              7/120 = 0.058333

                               

                              The error factor is 0.000294 is very low, for that reason, the factor 7:120 is used by EIGRP.

                               

                              Now, the amount of traffic send across a particular path is inversely proportional to the path’s metric, in other words: "lower metric, more packets", "higher metric, less packets"

                               

                              EIGRP unequal cost is a topic of CCNP Route.

                              *****EDITED*******

                              Best regards!

                              • 12. Re: EIGRP unequal load balancing
                                Ping

                                Ing_Percy wrote:

                                 

                                Hi!

                                 

                                The calculation of these values are made aproximatelly by maths

                                156160/2690560 = 0.058039

                                7/20 = 0.058333

                                 

                                The error factor is 0.000294 is very low, for that reason, the factor 7:20 is used by EIGRP.

                                 

                                Now, the amount of traffic send across a particular path is inversely proportional to the path’s metric, in other words: "lower metric, more packets", "higher metric, less packets"

                                 

                                EIGRP unequal cost is a topic of CCNP Route.

                                 

                                Best regards!

                                 

                                Hi Ing,

                                 

                                Your math is off

                                 

                                7 / 20 = 0.35 

                                 

                                7 / 120 = 0.058333

                                 

                                So the factor is 7 : 120.

                                • 13. Re: EIGRP unequal load balancing
                                  Ping

                                  Ing_Percy wrote:

                                   

                                  HI!

                                   

                                  No.

                                  The ratio of traffic over each path is 7:120. It means that for each 7 packets in the first link, it is sends 43 packets in the other link.

                                  In maths, the ratio is presented by: 7k/120k, where "k" is a constant integer greater than zero.

                                   

                                  Best regards!

                                   

                                  You got yourself confused.

                                   

                                  It's correct to say the ratio is 7:120 or 7k / 120k. But you also said incorrectly that "for each 7 packets in the first link, it sends 43 packets in the other link." The math just doesn't match up.

                                  • 14. Re: EIGRP unequal load balancing
                                    Ing_Percy

                                    Hi!

                                     

                                    In this document of Navneet (All credits to him) explains the nature of the calculation with accurancy about traffic share:

                                    EIGRP Traffic Share Ratio Calculation

                                     

                                    If you have more questions about it, clic in the link above and consult him

                                     

                                    Best regards!

                                    1 2 Previous Next