Two controllers have the same MAC address on Cisco Aironet 2800

[zeros66]

Hi everyone,

I'm running Cisco Aironet 2800 Mobility Express, I have two controllers installed on two AP 2800 devices, each managing 16 and 35 APs on two different VLANs in turn. However, today I noticed that both controllers are using the same MAC address. I also observed the following log message on the core switch: "Host abc.xyz in VLAN A is flapping between port Te2/0/3 and port Te2/0/37." (host abc.xyz is controller's MAC address)

The network seems normal, but when I ping the gateway via Wi-Fi, the latency seems abnormal, usually >1, sometimes very high (>100), and occasionally drops one line. When I connect to the network by wire, everything is normal, and the ping to the gateway is <=1.

Any help is appreciated!

[ammahend]

can you share “show interface detailed management” from both controller.

The flap message you would usually see if the client roams from one AP to another, that may be an expected behavior. 

About the ping delay, when did you start experiencing delay in ping? any changes ? if there is no performance issue I wont bother too much about it, comparing with Wired ping is not a fair comparison on a shared medium like wireless. 

[zeros66]

please see the attached for more details.

I'm not sure when the ping started behaving strangely—at least, for me—but it has been happening since yesterday. And yes, it affects the network system; sometimes it spikes suddenly, causing network interruptions. The latency to the gateway should be <=1ms, but mine is consistently at 3, 4, or even much higher.

[ammahend]

thats virtual mac address and virtual router ID is set to 1 (default) hence the MAC Address, This mac conflict may also be causing delay in ping. read section configuring vrid here 

[pieterh]

please verify the Mobility express controllers are individual controllers, and not setup as a redundant pair

if they where configured as redundant pair, then  they may now operate in  "dual-active" mode
that is, each controller serves a number of access points, but they do not communicate with each other  anymore
you risk losing configuration changes (saved on one, not copied to other)

[zeros66]

@pieterh no I have 2 virtual controller run in 2 VLANs, each managing 16 and 35 APs.

@ammahend  I changed the virtual MAC by changing VRID but seems it doesn't work, that could clear the warning message flapping on switch but the ping to the gateway still abnormal.

[ammahend]

has the ping ever worked properly ? was there a recent change or event before problem started? if you connect a laptop on same ip range as MEs and ping the gateway are you seeing different or same behavior ? have you tried a reload ? can you span the port where AP is connected to another port with wireshark run 100 ping again and share results ? was the arp table cleared on gateway after vrid change before you tried ping again ?