Problem with EIGRP path selection.

[ddebrova]

Hello everyone,

Can anyone explain why the Route paths are sorted in the following order even though the last link has a lower metric value than the second one?

Output from Router :

NC200-Core-1# sh ip eigr topology 10.20.10.64/26

IP-EIGRP (AS 210): Topology entry for 10.20.10.64/26
State is Passive, Query origin flag is 1, 1 Successor(s), FD is 3328
Routing Descriptor Blocks:
192.168.26.77 (Ethernet1/2), from 192.168.26.77, Send flag is 0x0
Composite metric is (3328/3072), Route is Internal
Vector metric:
Minimum bandwidth is 1000000 Kbit
Total delay is 30 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2
Internal tag is 0
192.168.26.9 (Ethernet1/1), from 192.168.26.9, Send flag is 0x0
Composite metric is (29212/3072), Route is Internal
Vector metric:
Minimum bandwidth is 90001 Kbit
Total delay is 30 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2
Internal tag is 0
192.168.26.37 (Ethernet1/40), from 192.168.26.37, Send flag is 0x0
Composite metric is (26368/26112), Route is Internal
Vector metric:
Minimum bandwidth is 100000 Kbit
Total delay is 30 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 2
Internal tag is 0

Thank you.

[Meheretab Mengistu · Accepted solution]

Hi,

Can anyone explain why the Route paths are sorted in the following order even though the last link has a lower metric value than the second one?

On the last link, RD > FD (26112 > 3328). Since the FD is 3328 and the link through Ethernet1/40 has higher RD (26112), it does not meet the Feasibility condition (which states the RD < FD to be a feasible successor.)

HTH,

Meheretab

[RyanB]

Actually, while that is true ( about the feasibility condition ) that is not why they are listed in that particular order. For example:

EIGRP-IPv4 Topology Entry for AS(1)/ID(10.0.14.1) for 192.168.0.0/24
  State is Passive, Query origin flag is 1, 1 Successor(s), FD is 3328
  Descriptor Blocks:
  10.0.12.2 (GigabitEthernet0/0), from 10.0.12.2, Send flag is 0x0
      Composite metric is (3328/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 1000000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.2
  10.0.13.2 (GigabitEthernet1/0), from 10.0.13.2, Send flag is 0x0
      Composite metric is (29184/2816), route is Internal
      Vector metric:
        Minimum bandwidth is 90001 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.3
  10.0.14.2 (GigabitEthernet2/0), from 10.0.14.2, Send flag is 0x0
      Composite metric is (26368/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 100000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.4

The RD through 10.0.13.2 and 10.0.14.2 are BOTH less than the FD through 10.0.12.2, making them feasible successors. Given that, the order is still the same in that the FD of 26368 is listed AFTER the FD of 29184.

Summarized here:

P 192.168.0.0/24, 1 successors, FD is 3328
        via 10.0.12.2 (3328/3072), GigabitEthernet0/0
        via 10.0.13.2 (29184/2816), GigabitEthernet1/0
        via 10.0.14.2 (26368/3072), GigabitEthernet2/0

I would speculate that the order is successor first, then the lowest reported distances in order.

Actually, to soon to make that assumption.

[RyanB · Accepted solution]

Did some quick testing, I think it has mostly to do with the order or learning/processing the updates.

The successor route is for sure listed first.

Here is sample output to explain why I believe it has to do with learning/processing order. I had a EIGRP process running, with routes learned. Shutdown the process, started it back up again, and the order of entries change (with the exception of the successor).

EIGRP-IPv4 Topology Entry for AS(1)/ID(10.0.14.1) for 192.168.0.0/24
  State is Passive, Query origin flag is 1, 1 Successor(s), FD is 3072
  Descriptor Blocks:
  10.0.15.2 (GigabitEthernet3/0), from 10.0.15.2, Send flag is 0x0
      Composite metric is (3072/2816), route is Internal
      Vector metric:
        Minimum bandwidth is 1000000 Kbit
        Total delay is 20 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.5
  10.0.12.2 (GigabitEthernet0/0), from 10.0.12.2, Send flag is 0x0
      Composite metric is (3328/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 1000000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.2
  10.0.14.2 (GigabitEthernet2/0), from 10.0.14.2, Send flag is 0x0
      Composite metric is (26368/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 100000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.4
  10.0.13.2 (GigabitEthernet1/0), from 10.0.13.2, Send flag is 0x0
      Composite metric is (29184/2816), route is Internal
      Vector metric:
        Minimum bandwidth is 90001 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.3

Order is:

- Successor (R5 @ 10.0.15.2)

- R2 @ 10.0.12.2

- R4 @ 10.0.14.2

- R3 @ 10.0.13.2

R1(config)#router eigrp 1
R1(config-router)#shut

*Nov 10 18:47:09.291: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.12.2 (GigabitEthernet0/0) is down: shutdown
*Nov 10 18:47:09.291: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.15.2 (GigabitEthernet3/0) is down: shutdown
*Nov 10 18:47:09.291: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.13.2 (GigabitEthernet1/0) is down: shutdown
*Nov 10 18:47:09.291: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.14.2 (GigabitEthernet2/0) is down: shutdown

R1(config-router)#no shut

*Nov 10 18:47:23.003: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.15.2 (GigabitEthernet3/0) is up: new adjacency
*Nov 10 18:47:23.131: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.12.2 (GigabitEthernet0/0) is up: new adjacency
*Nov 10 18:47:23.499: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.14.2 (GigabitEthernet2/0) is up: new adjacency
*Nov 10 18:47:26.519: %DUAL-5-NBRCHANGE: EIGRP-IPv4 1: Neighbor 10.0.13.2 (GigabitEthernet1/0) is up: new adjacency

Shutdown / No shutdown the process, and the new results are...

EIGRP-IPv4 Topology Entry for AS(1)/ID(10.0.14.1) for 192.168.0.0/24
  State is Passive, Query origin flag is 1, 1 Successor(s), FD is 3072
  Descriptor Blocks:
  10.0.15.2 (GigabitEthernet3/0), from 10.0.15.2, Send flag is 0x0
      Composite metric is (3072/2816), route is Internal
      Vector metric:
        Minimum bandwidth is 1000000 Kbit
        Total delay is 20 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.5
  10.0.13.2 (GigabitEthernet1/0), from 10.0.13.2, Send flag is 0x0
      Composite metric is (29184/2816), route is Internal
      Vector metric:
        Minimum bandwidth is 90001 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.3
  10.0.14.2 (GigabitEthernet2/0), from 10.0.14.2, Send flag is 0x0
      Composite metric is (26368/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 100000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.4
  10.0.12.2 (GigabitEthernet0/0), from 10.0.12.2, Send flag is 0x0
      Composite metric is (3328/3072), route is Internal
      Vector metric:
        Minimum bandwidth is 1000000 Kbit
        Total delay is 30 microseconds
        Reliability is 255/255
        Load is 1/255
        Minimum MTU is 1500
        Hop count is 1
        Originating router is 192.168.0.2

- Successor (R5 @ 10.0.15.2)

- R3 @ 10.0.13.2

- R4 @ 10.0.14.2

- R2 @ 10.0.12.2

That being said, I wouldnt worry too much about the order in which they are listed. Although, it would be a little more convienient if there was a particular order, such as advertising router-id or feasible distance.

[ddebrova]

Meheretab and Ryanb ,

Thank you for your replies.

I am worried if the successor link drops will it failover to the second listed link with a higher metric or will it prefer the last link in the table with a lower metric. I want it to failover to the last one as that is the peer link to the other Nexus Core router. Since it is in production I will have to schedule and test.

@Meheretab Mengistu   I thought if  RD>FD that link shouldn't appear in the topology table at all, correct me if I am wrong.

[RyanB · Accepted solution]

The Successor will fail over to the feasible successor:

192.168.26.9 (Ethernet1/1), from 192.168.26.9, Send flag is 0x0
  Composite metric is (29212/3072), Route is Internal
  Vector metric:
  Minimum bandwidth is 90001 Kbit
  Total delay is 30 microseconds
  Reliability is 255/255
  Load is 1/255
  Minimum MTU is 1500
  Hop count is 2
  Internal tag is 0

If you run 'show ip eigrp topology', the 10.20.10.64/26 network should look like this:

P 10.20.10.64/26, 1 successors, FD is 3328
        via 192.168.26.77 (3328/3072), Ethernet1/2
        via 192.168.26.9 (29212/3072), Ethernet1/1

192.168.26.37 will not appear in this output because its RD (26112) is higher than the FD of the successor (3328).

Let me know if the output is different, im curious.

[Meheretab Mengistu]

@Meheretab Mengistu   I thought if  RD>FD that link shouldn't appear in the topology table at all, correct me if I am wrong.

Actually, when you run 'sh ip eigrp topology all-links', you will see successor, feasible-successor and non-successor routes. So, even if RD>FD, you will see it if you are looking for all-links related with the particular route (it is received from a neighbor). However, you will not see non-successor routes (which are not successor nor feasible-successor) when you run 'sh ip eigrp topology'.

HTH,

Meheretab

[Julio E. Moisa]

Hi

The first path is your succesor, basically it has the lowest FD metric to the destination. 

The second path meets the Feasible condition, where it mentions: the reported/advertised distance must be less of the Succesor Feasible Distance.

The third path should not be included on the EIGRP's topology because the reported distance is higher than the Succesor FD.  26112 is higher than 3328. So it will not be considered as a Feasible succesor and if the path 1 and 2 are unavailable, the router will make a difussing computation process to find a new path. 

You can execute a: clear ip eigrp neighbor in order to refresh the table (if it is on production you must do it after business hours).

Hope it is useful

:-)

[paul driver]

Hello

The FD (feasible Distance) is the single best record for that particular destination (10.20.10.64/26) calculated through that specified neighbor (in this case 192.168.26.77) since that that prefix (destination) went from a active to passive state.

Please review this ( the most definitive explanation on dual computation)  by Mr Peter Paluch

res
Paul

[ddebrova]

Thank you, everyone, for taking time to look at this.

I think I know what to do next.