EIGRP or CEF load-sharing question

[Ryan YC]

Hi,

My company is having a routing or load-sharing problem so I hope you can help me. The following diagram is the current situation:

I don't know when the direct layer 2 connection between SW1 and SW2 went down, the link is a 20km fiber connection between two buildings so we can't fix it quickly, and we don't have a backup connection between the switch so HSRP has become Active-Active. 

We are using EIGRP equal-cost load sharing in the environment. but it seems load-sharing is the reason to make me having this problem. When I tried to connect to the file server from the Core-1, I won't be able to reach the target, because Core-2 sent my traffic to the incorrect segment which was not connecting to the file server. if I shut down the connection between Core-2 and SW2, I can connect to the server again.

We currently use several static routes to specify the possible route, but we would like to think about a global solution for now and in the future, we wouldn't like to prioritize the EIGRP metric or filter the routes at this moment because there are actually over hundreds of servers connected behind the SW3 or SW4. The above diagram is just one of the small segments in our network.  

I have thought there maybe are some technologies that I can let the Core-2 automatically determine which route is possible to reach the target devices. I have been reading the articles of CEF and EIGRP and trying to help myself.

Thank you!

[Ryan YC]

Thank you for helping!

Unfortunately, we cannot remove the connection and connect between SW3 and SW4, as the switches are layer 2 switches (Cisco IE3000 series), and due to the distance issue, we cannot make the engineering very easy. 

The VLAN Interfaces on the "SW1" and "SW2" are still working because there are actually more devices connected plus the layer 2 trunks are still working as well.

[David Ruess]

Hello,

As @Richard Burts pointed out and I agree that once the L2 link between S1 and S2 are broken then you have the same L3 network in 2 different places without any other connection between them. So EIGRP thinks it can get to the network segment either way (from your description). You may need to look at modifying your network structure to fix some of your issues. In the interim you could put a Policy Based Routing statement that says anything destined for the file server take this route no matter what. This would be a temporary fix while you figure out how to redesign your network.

Also to see if EIGRP is load balancing you can issue the show ip eigrp topology all-links command. This will show if there are multiple paths for the same network. 

Hope that helps

-David

[Joseph W. Doherty]

As other posters have already noted, especially Rick, your topology information is unclear/incomplete.

For example, you show multiple gateway IPs on switches 1 and 2, but the diagram symbols for those switches, both are just L2 switches.(?)  You also show HSRP running on both, which also would normally only be present on L3 switches.  You further, mention the down link between those two switches is only L2, but don't note whether it's a L2 trunk. You also mention EIGRP ECMP, which would seem to make sense if applied to Core-2 and if switches 1 and 2 were L3.

If switches 1 and 2 are L2 facing switches 3 and 4, and have/had a L2 trunk between them, losing the trunk would partition your networks, fouling up routing (regardless of IGP, static or CEF) and making HSRP active on both switches 1 and 2 (as you describe).

The "fix" is to trunk all your VLANs, from switches 1 and 2, to Core-2.  Assuming those links a p2p now, you could redefine those p2p to each have be on its own (dedicated) VLAN, and continue to treat it at p2p links.  If and when you fix the "down" cross link, you just need to insure some variant of STP is running to block one of the (then) circular links.

What I just described, might be the fastest and way to restore correct operation with minimal changes.  However, I would suggest considering a "better" topology design (to take better advantage of switches 1 and 2, assuming they are L3 switches.