cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
16733
Views
5
Helpful
3
Replies

CDP-4-NATIVE_VLAN_MISMATCH

Jeroen Huysmans
Level 1
Level 1

Hi,

I have some weird CDP issue reporting a mismatched native vlan on a trunk:

SWITCH-A:

%CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on GigabitEthernetx/xx (999), with SWITCH-B GigabitEthernety/yy (516)

SWITCH-B:

%CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on GigabitEthernety/yy (516), with SWITCH-A GigabitEthernetx/xx (999).

both switch have etherchannel holding 2 gigabit interfaces, CDP msg'es come for every port in channel.

config SWITCH-A (WS-C6509-E,122-33.SXH4):

vlan dot1q tag native

interface Port-channel2

description *** SWITCH-B po1 ***

switchport

switchport trunk encapsulation dot1q

switchport trunk native vlan 516

no switchport trunk native vlan tag

switchport trunk allowed vlan 1,516 (and others)

switchport mode trunk

switchport nonegotiate

config SWITCH-B (OS-CIGESM-18TT-EBU, 121-22.EA12):

interface Port-channel1

description *** SWITCH-A po2 ***

switchport trunk native vlan 516

switchport mode trunk

switchport nonegotiate

speed 1000

duplex full

flowcontrol send off

Reason for NOT tagging native VLAN on SWITCH-A: apparently there is an issue with tagged frames & PXE-boot.

I seems SWITCH-A is sending the wrong CDP info (as it states SWITCH-A has native VLAN 999)

checked switchport on SWITCH-A:

Trunking Native Mode VLAN: 516

Administrative Native VLAN tagging: disabled

Operational Native VLAN tagging: disabled

so, SWITCH-A is using native VLAN516 (as configured).

Why does CDP keep sending me this info about mismatched native vlan's? Disabling CDP stops the messages, but that's not a solution...

I have a similar setup (same type of switches, same IOS), without CDP errors...

jeroen

3 Replies 3

Peter Paluch
Cisco Employee
Cisco Employee

Jeroen,

This is interesting. Can you with 100% certainty confirm that these two switches are really connected to each other without any intermediary device inbetween? Is it by any means possible that you are in fact connected to a different switch? Is it possible to physically inspect the cabling?

By the way, what is the VLAN 999? Is it defined? It seems to be used by the C6509.

Also please inspect the configuration of the physical interfaces under the port channels - perhaps there is a configuration mismatch between the setting of the physical ports and the setting of the port-channels.

Best regards,

Peter

Peter,

thanks for your reply.

I'm 100% sure these 2 switches are directly connected to each other (2 gigabit interfaces into EtherChannel).

VLAN 999 used to be my native vlan, I still use it on trunks between my 6500 switches. I had to use VLAN516 as native vlan when connecting to the IBM BladeCenter stacks (PXE-boot).

physical interfaces are configured with native vlan516.

interface GigabitEthernet4/25

description *** SWITCH-B 0/17 ***

switchport

switchport trunk encapsulation dot1q

switchport trunk native vlan 516

no switchport trunk native vlan tag

switchport trunk allowed vlan 1,516 (and others)

switchport mode trunk

switchport nonegotiate

logging event link-status

speed 1000

duplex full

channel-group 2 mode on

spanning-tree portfast

end

interface GigabitEthernet7/25

description *** SWITCH-B 0/18 ***

switchport

switchport trunk encapsulation dot1q

switchport trunk native vlan 516

no switchport trunk native vlan tag

switchport trunk allowed vlan 1,516 (and others)

switchport mode trunk

switchport nonegotiate

logging event link-status

speed 1000

duplex full

channel-group 2 mode on

spanning-tree portfast

end

interface Port-channel2

description *** SWITCH-B Po1 ***

switchport

switchport trunk encapsulation dot1q

switchport trunk native vlan 516

no switchport trunk native vlan tag

switchport trunk allowed vlan 1,516 (and others)

switchport mode trunk

switchport nonegotiate

I tried stopping CDP (no cdp run) and re-enabling it, but the issue remains. It keeps claiming switch A is using native vlan999 while switch B is in vlan516.

jeroen

interface GigabitEthernet4/25
description *** SWITCH-B 0/17 ***
switchport
switchport trunk encapsulation dot1q
switchport trunk native vlan 516
no switchport trunk native vlan tag
switchport trunk allowed vlan 1,516 (and others)
switchport mode trunk
switchport nonegotiate
logging event link-status
speed 1000
duplex full
channel-group 2 mode on
spanning-tree portfast
end

interface GigabitEthernet7/25
description *** SWITCH-B 0/18 ***
switchport
switchport trunk encapsulation dot1q
switchport trunk native vlan 516
no switchport trunk native vlan tag
switchport trunk allowed vlan 1,516 (and others)
switchport mode trunk
switchport nonegotiate
logging event link-status
speed 1000
duplex full
channel-group 2 mode on
spanning-tree portfast
end

interface Port-channel2
description *** SWITCH-B Po1 ***
switchport
switchport trunk encapsulation dot1q
switchport trunk native vlan 516
no switchport trunk native vlan tag
switchport trunk allowed vlan 1,516 (and others)
switchport mode trunk
switchport nonegotiate

Hello Jeroen,

Check if you are running same CDP version between the two switches. if not try to sync it up.

CDP version 2 has a Time Length Variable (TLV) field that carries native VLAN information. CDP version 1 does not. If a switch running version 2 receives a frame from a switch running version 1, it sees no native VLAN information and assumes the native VLAN is a zero. As its native VLAN would be a non-zero value, it does not generate this informational message. Issue the show cdp command to check the CDP version running on both devices.

If you are running the same version and still you have the problem. you are possible hitting the bug CSCtc12295 .(WIP)

Try upgrading it to SXJ and see if its stops.

Thanks,

Ricky Micky

*Pls rate useful posts

Review Cisco Networking for a $25 gift card