SG 300 Duplicate IP address issue

[jbritcliffe1]

Hi,

I've searched through the forums and done some reading but have not been able to resolve a duplicate IP address error that I'm getting on numerous SG 300-10's.

The message I'm seeing is:

10-Jun-2014 11:25:57 :%IPADTBL-N-IPDUPLICATE: Duplicate IP address 192.168.20.223 from MAC 68:86:a7:4f:43:18 was detected on VLAN 6, port gi9, aggregated (3)

10-Jun-2014 11:25:01 :%IPADTBL-N-IPDUPLICATE: Duplicate IP address 192.168.20.223 from MAC 68:86:a7:4f:4d:99 was detected on VLAN 6, port gi9, aggregated (3)

Both MAC addresses are trunk ports on 2960's  I've compared the config of the 2 2960's with others that we have in production and cannot see anything different (excluding the obvious IP / name).  The SG 300's and 2960's all uplink to our 4500 core switches.

The duplicate IP address in the message is always the IP address of the SG 300 that the message appears on, and I do not see any messages on the 2960's.

I'm just wondering if anyone has any pointers of where I should be looking next.....

Thanks

[Tom Watts]

Are you running PVST on the 2960's? Try changing the spanning tree mode to RSTP without PVST if it's not already. The SG300 uses only one system MAC ID. The 2960 can use port MAC, if you have 2 links between the switches that may cause confusion.

To troubleshoot, I'd make everything as simple as possible;

 -2960 running with nothing else connected

 -SG300 running with nothing else connected

 -Move 1 wire between 2960 and SG300

What's the result?

Depending on the result, then start SLOWLY normalizing your network until the error shows up.

[chrebert]

Hello,

This message is a bit of a bug, it has to do with the IP device tracking feature in IOS.

From what I have read on escalated cases it is actually a bug in IOS.  The workaround is to disable IP device tracking on the interface that connects to the SGs.  I am not that familiar with that particular IOS feature, but I have read it has something to do with Gratuitous ARP as well.

All of the cases I have read resolve this issue by disabling that feature on the Catalyst switch, which can usually be done on a port-by-port basis, so you should still be able to leave it on everywhere else if it is required.

I wish I could give you a bug number or something, but it is more of an issue with the Catalyst, so I don't have as many details on that as I would with an SBSC product.

Just for my curiosity, is it actually causing an issue on the network or is it just a log message? 

Hope that helps and thank you for choosing Cisco,

Christopher Ebert - Advanced Network Support Engineer

Cisco Small Business Support Center

*please rate helpful posts*

[jbritcliffe1]

Hi there,

No we are not seeing any issues on our network, just the logs on the SG-300's.

It's interesting that while all our 2960's have the same base configuration it's only a couple of them that are being reported on the SG-300's

Thanks

[ciscokid6969]

I am seeing a very similar issue on a dell 5548P switch. I traced the mac address to the interface on a new 2960x which is connected to another switch (dell 6224) that they are both connected to. It is not configured to have that IP address, so like you said I think its an informational bug.

The issue occurs on the 5548 connected like:

5548 ---2 ch LAG --->6224 ---trunk--->2960x

It seems to be just informational. The ip address gets me to the correct machine, so i doubt its an actual conflict of address, however it is weird!