Native VLAN Mismatch between 2 internal switches

[BillyBobKnob]

I have this error showing on a member switch of a cluster setup through the web interface.

I have verified that both switches are set to use VLAN1 as the native vlan, and since these are internal switches CDP is running on both.

Both are Catalyst2924XLs. The member switch is running IOS v11.2 and the command switch is running IOS v12. Both switches are also running CDPv2.

Any ideas on how to stop this error message ?

Thanks

[hmkoh]

Just make sure that the switchport connecting both the switches need to be native vlan 1.

[BillyBobKnob]

I did show vlan on both switches and all the ports on both are part of the default VLAN 1.

[BillyBobKnob]

Let me start over and describe the setup without pasting all my configuration settings here.

I have a core switch A (Catalyst 2950, IOS v12.1) that attaches to another core switch B (2950, v12.1) at a backup site through our routers and a private line.

These core switches are connected through a VLAN other than the default VLAN1.

Then the core switchA is connected to a cluster root switchC (2924XL,IOSv12.0) which has 2924XL member switches D(IOSv11.2),E(v12.0),F(v12.0) and G(v12.0). This cluster root was setup through the web interface so it added the normal cluster member and nat statements automatically to make it the cluster root.

The native VLAN mismatch error was first appearing on C so on A trunking was turned off (switch mode access) since the 2950's auto negotiate trunking and CDP was disabled as well.

This took care of the VLAN mismatch warning error.

Now we have the VLAN mismatch warning on D.

There is no VLAN other than the default of 1 between C and D and CDP is runnning on both as it is needed.

Could it be that D is running IOS v11.2 as there is just the basic setup on C and D ?

Hope this makes it clearer.

Thanks

[ganganathshanmugalingam]

First of all I would like to tell you that this problem can be solved.

do you have a vtp domain on your network.

If so, please concider whter you need it or not.

You can manage your network wothout VTP domain easily.

If you dont have a VTP domain. Change the Switches configuration to

" vtp mode transparent"

then you will not get any VLAN mismatch unless you are trunking between the switches.

you said that at first you got a VLAN mismatch between A and C.

You sould not have turned trunking off.

Of course for trunking, both end interfaces should have the same NATIVE VLAN.

fortunately you can change the native VLAN for that interface only.

To trobleshoot the problem undo any changes you have made after the problem arise.

then troubleshoot the problem by equaling the Native VLAN for the trunk interface only.

[milan.kulik]

Hi,

AFAIK, this problem is caused by CDP ver.2 advertising "native VLAN" field even on access ports.

I've seen the "native VLAN mismatch" syslog error message when two switches were connected via access ports assigned to a different VLANs.

The cure was one of following:

a) put both ports to the same VLAN

b) disable CDP ver 2 or CDP completely

c) set logging level cdp 3 default on CatOS switch

You can check the native VLAN assignment via issuing:

sh cdp neigh detail

and see if both ports involved are assigned to the VLAN1.

I'm not sure if VTP transparent mode would stop the error messages, I'd guess no. And it's more comfortable using VTP in your network then configure all VLANs manually on each switch.

Regards,

Milan