cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1584
Views
11
Helpful
6
Replies

EIGRP Problem on ASR 1002

Julio Carvajal
VIP Alumni
VIP Alumni

Hello Experts,

I have a pair of ASR 1002 running version IOS XE Version: 03.09.01.S  15.3(2)S1 that are showing a completely weird issue.

Here is the Scenario:

                          DATA CENTER In San Jose

                             |                            |

                             |                            |

                             _____________              

                        ASR 1                     ASR2

                              _____________

So basically

Data-Center in San Jose connects via 2 GRE tunnels (to each ASR) on a remote location. They are  running EIGRP over theGRE  link.

ASR1 and ASR 2 have 2  links between each other (redundancy purposes).

Now the problem:

Everything is good as long as we have only one link between the ASR enable.

As soon as we enable EIGRP over the the secondary link between each other the CPU starts increasing  (because of a Stuck In Active Routes).

I started to check why this happen and then I found that :

ASR1 installs in it's routing table the networks for the data center via ASR2 instead of the Tunnel interface. This means that is not taken in consideration the best route (Lowest EIGRP Metric) That's right!! Does not make any sense right? I mean why would the router be using the less optimal path????

So here is what I have:

  • Both routes are in the EIGRP topology table ( Again the one installed in the routing table is the one with the worst metric )
  • EIGRP relationship is built across all of the interfaces ( the 3 reference here) and the CPU increases
  • If we run EIGRP over 2 links, we are good.
  • Both ASRs run the same IOS code version and no filtering is being done.

I cannot post any configuration (security purposes) so I know that it's almost impossible to troubleshoot So I just want to know if you are aware of any bug or what should I look at for this kind of issues. For me it sounds like a bug as the configuration is perfect and I have the same deployment on multiple sites.

What could be causing the EIGRP route not being installed on the Routing table?

  • Routers-ID are different
  • Both are seen on the EIGRP topology database.

If I fix the issue of the EIGRP not taken into consideration the tunnel interface routes then EIGRP can be enabled over all of the links and the CPU will not increase.

I would appreciate any help on this,

Regards,

Julio Carvajal

Julio Carvajal
Senior Network Security and Core Specialist
CCIE #42930, 2xCCNP, JNCIP-SEC
6 Replies 6

Vishesh Verma
Level 1
Level 1

Hi Julio,

ASR-1 should have 3 routes in EIGRP topology table, 1 via tunnel, 1 via primary link b/w ASRs and 1 via secondary link b/w ASR. The one installed in the routing table is via primary link b/w ASR, even though it has higher metric than the 1 received from tunnel?

My Questions,

Is the route via tunnel is Active? Do you have any distance command configured for eigrp routes received from primary link?

- Vishesh

Hello,

Is the route via tunnel is Active?

No, without the 3 link between each other, We have 2 routes (One from the Tunnel and one from the Neighbor ASR) in the passive state. Only that the installed one is the less preferred.

Remember that the real issue is why while having 2 links being used, we preffered the less optimal route.

Do you have any distance command configured for eigrp routes received from primary link?

No, AD for both of them are the same. No distance Command changing the AD.

Regards,

Julio Carvajal
Senior Network Security and Core Specialist
CCIE #42930, 2xCCNP, JNCIP-SEC

Hi Julio

Rather than running blindly on this case, I would suggest you to open a TAC case, if you have not done that already.

-Vishesh

I would have done that on the first place instead of comming here, that's not an option right now

Anyone else, I would appreciate any comments

Julio Carvajal
Senior Network Security and Core Specialist
CCIE #42930, 2xCCNP, JNCIP-SEC

Julio,

Is it possible to at least post the output of the show ip eigrp topology that would show the route taking the wrong path? At least something, perhaps addresses changed to harmless private space?

I was just thinking if it is possible that there is some kind of typo in the configuration resulting in the interconnected devices not using the same netmask on a common network. That could, under circumstances, lead to similar symptoms: a router learning about a network it is supposed to be directly connected to via a routing protocol from another router that advertises the network correctly.

Best regards,

Peter

ashirkar
Level 7
Level 7

Hi Julio,

Hope you are doing well,

You can remove all confidential information and post your config(Example: remove public IP if any & replace them with private IP address) and  post output of taking specific route like "

sh ip eigrp topology x.x.x.x" so that community users can help you.It will blind troubleshooting for us unless someone has gone through same scenario/issue.

Regards,

Ashish

TCM-NI

Review Cisco Networking for a $25 gift card