cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1049
Views
5
Helpful
15
Replies

Having issues with VLAN on 3548 switch. Please help

j.zortman
Level 1
Level 1

Hello,

I am running a VLAN on a switch it runs fine until I plug a certain machine into it I get the following erorr:

2y43w: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on FastEthernet0/42 (10), with sw-2 FastEthernet0/45 (1).

Interface 45 is the one causing the issue. When I put interface 45 on the VLAN it is supposed to be which is VLAN10 it freezes up our network and starts relearning addresses

2y43w: %RTD-1-ADDR_FLAP: FastEthernet0/45 relearning 6 addrs per min

Any ideas?

Let me know if you need more info.

Thanks!

15 Replies 15

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello Justin,

you have a Native Vlan mismatch:

the two sides don't agree on the native vlan number that is the vlan whose frames are sent out the 802.1Q trunk untagged = as normal ethernet frames with no vlan tag in them

verify by comparing config

sw1# sh run int f0/42

compare with

sw2# sh run int f0/45

the first one has a line

switchport trunk native vlan 10

that is missing on the second (vlan 1 native is default )

fix this mismatch or you can have IP connectivity problems because each switch thinks untagged frames belong to a different broadcast domain

Hope to help

Giuseppe

Thanks for the reply!

Right it does. However the problem is, that both 42 and 45 should be in VLAN 10. When I try to add 45 to VLAN I get some sort of loop where all the lights flash like crazy on the switch and I get the following message:

2y43w: %RTD-1-ADDR_FLAP: FastEthernet0/45 relearning 6 addrs per min

It basically shuts down a part of our network when this happens.

Hello Justin,

you need to hardcode they are access ports

sw1 # int f0/42

switchport

switchport mode access

switchport access vlan 10

the same on other side

they can be trying to negotiate a trunk dynamically and so then you have unexpected behaviuor

Hope to help

Giuseppe

Giuseppe, I have them both setup that way.

Thanks.

It looks like you have a physical wiring loop from one f0/42 to f0/45 and you will have to figure out where that is . If you have portfast on the ports turn it off temporarily while you plug those ports in and see if spanning tree will put one of the ports into blocking mode . Probably someone has one of those nice little home switches at their desk and has managed to loop the wiring going back to the switch , see it all the time .

Hello Glen,

you're right they see each other CDP messages

Best Regards

Giuseppe

Thanks Glen.

Both ports 42 and 45 are servers and there is nothing but servers on this switch.

What should I try?

Thanks!

I would trace out the wiring on 42 and 45 and see where they go . Someone may have done some creative networking that you don't know about .

They both go to the servers.

Now would this make a difference, these are 2 servers that have 2 other virtual machines on them and those have virtual interfaces.

Hmm So I unplugged fa0/42 and put fa0/45 and I dont get the errors. So there is something between those 2 machines that they dont like and it creates some sort of loop.

Is there a way to test and get to the bottom of this?

Thanks.

I turned debug ethernet-controller address on and I gte the following messages...any ideas?

2y43w: Add address 0016.32, on port Fa0/42 vlan 10

2y43w: Add address 0016.32, on port Fa0/45 vlan 1

2y43w: Delete address 0019.b, on port Fa0/42 vlan 10

2y43w: Add address 0014.fd, on port Fa0/45 vlan 1

2y43w: Add address 0006.5a, on port Fa0/45 vlan 1

2y43w: Add address 0006.5a, on port Fa0/42 vlan 10

2y43w: Add address 000d.3, on port Fa0/45 vlan 1

2y43w: Add address 000d., on port Fa0/42 vlan 10

Hello Justin,

I think the servers are bridging to each other and CDP messages are L2 multicast so they pass through.

You should verify the networking config of the servers both physical and virtual instances.

Hope to help

Giuseppe

Yes I think it is something with the servers as well. We are running Zend Core on there with it having 2 other virtual machines on it.

Just cant seem to figure out where the issue is on the interfaces.

Hello Justin,

the servers have at least another NIC each of them and they are bridging between external NIC and internal NIC and then the second from internal NIC to external NIC

where I called external NICs the ones that are cabled to your switches' ports.

on www.cisco.com/go/srnd there is a design guide for VMware.

However, if there are "servers" people I would involve them in this issue: explain them that the two servers are disturbing the infrastructure networking and you should disable the ports for the network safety ...

Hope to help

Giuseppe

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: