Two routers are losing eigrp adjacency

[sysadmin@cwcmh.org]

I have my data center router and 10 remote routers that all connect back to the data center in a hub and spoke topology. I use an ISP to connect the remote sites to the data center using MPLS, they say it's a theoretical layer 2 connection. One of my remote routers is unable to keep it's eigrp adjacency with that data center router. These are the logs i get,

remoterouter#show loggi | i DUAL-5-NBRCHANGE
216491: *Oct 2 05:42:29.414: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is up: new adjacency
216812: *Oct 2 05:43:48.928: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is down: retry limit exceeded
216828: *Oct 2 05:43:52.346: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is up: new adjacency
217147: *Oct 2 05:45:11.857: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is down: retry limit exceeded
217165: *Oct 2 05:45:15.627: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is up: new adjacency
217486: *Oct 2 05:46:35.140: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is down: retry limit exceeded
217492: *Oct 2 05:46:35.325: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is up: new adjacency
217813: *Oct 2 05:47:54.838: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is down: retry limit exceeded

remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.254.27.1 Gi0/0/0.1027 13 00:01:26 1 5000 1 11239
remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.254.27.1 Gi0/0/0.1027 12 00:01:28 1 5000 1 11239
remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)

remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.254.27.1 Gi0/0/0.1027 14 00:00:00 1 2000 1 0
remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.254.27.1 Gi0/0/0.1027 13 00:00:01 1 2000 1 0
remoterouter#show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(33)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.254.27.1 Gi0/0/0.1027 12 00:00:02 1 3000 1 0

The isp says its my network, and CIsco tac says it's the ISP. Not sure where to go from here.

[MHM Cisco World]

Try use under tunnel

Ip eigrp bandwidth percentage 50

And check 

MHM

[David Ruess]

Hello,

So not quite sure where the issue is just yet. Can you provide EIGRP and interface configuration being used on both devices?

So from the logs it brings up the Adjaceny because it hears the remote end trying to establish a neighborship. It then sends out its own EIGRP messages to confirm, but these messages never reach the other end. So maybe multicast is not working trough the ISP on the return path or is being limited. Could be an ACL blocking EIGRP or other protocols limiting the exchange of routing information.

-David

[Giuseppe Larosa]

Hello sysadmin@cwcmh.org ,

>> 216491: *Oct 2 05:42:29.414: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is up: new adjacency
216812: *Oct 2 05:43:48.928: %DUAL-5-NBRCHANGE: EIGRP-IPv4 33: Neighbor 10.254.27.1 (GigabitEthernet0/0/0.1027) is down: retry limit exceeded

EIGRP hellos are OK but EIGRP update packet is not received or not acknowledeged as a result of this the local node sends the update packet again and  again up to N times.

then it says it is too much and the neighbor state machine is reset and it starts again.

Check the MTU.

if using authentication try to remove it

OR :

the layer 2 has troubles with multicast try to use unicast neighbors command under router eigrp-

Sometimes this happens and the fix it to move to use unicast packets only both for hellos and for updates.

Hope to help

Giuseppe

[sysadmin@cwcmh.org]

Thank you everyone. The issue is resolved. It was the fault of my isp. They still haven't pinpointed why a connection can't be made. What they did was swap my wan link port to another open port on their router. My remote router immediately came online afterwards.. @Giuseppe Larosa @David Ruess @MHM Cisco World 

[MHM Cisco World]

Thanks a lot for update us

MHM