VTP domain migration

[wim_depauw]

Hi,

we're having a problem with the migration of VTP domaing .

For the moment our server switches are vtp server in domain X . Our distribution switches are vtp client in domain X and the access switches are in transparent mode .

The goal is to make our distribution switches VTP server and access switches vtp client .

We already did this successfully for several distrib switches and now we're trying to migrate the main building .( about 21 access switches )

When configuring the distrib switch as vtp server we loose our connection with the access switches and it doesn't come up again . In the past we had a outage of about 30 sec for spanning-tree .

We get the following message on our access switches

Aug 11 16:41:53.182 CEST: %SPANTREE-7-RECV_1Q_NON_TRUNK: Received 802.1Q BPDU on non trunk FastEthernet0/3 VLAN20.

Aug 11 16:41:53.182 CEST: %SPANTREE-7-BLOCK_PORT_TYPE: Blocking FastEthernet0/3 on VLAN0020. Inconsistent port type.

Aug 11 16:42:11.282 CEST: %SPANTREE-7-RECV_1Q_NON_TRUNK: Received 802.1Q BPDU on non trunk FastEthernet0/4 VLAN20.

Aug 11 16:42:11.282 CEST: %SPANTREE-7-BLOCK_PORT_TYPE: Blocking FastEthernet0/4 on VLAN0020. Inconsistent port type.

Aug 11 16:43:38.381 CEST: %LINK-3-UPDOWN: Interface FastEthernet0/4, changed state to down

on the website it says that link should be restored automatically .

Our ports are configured as trunks with native vlans defined

switchport access vlan 20

switchport trunk encapsulation dot1q

switchport trunk native vlan 20

switchport trunk allowed vlan 1,20,700-799

no ip address

duplex full

Does anybody have an idea about this ? My guess is that it gets stuck with recalculating the spanning-tree .I checked the bugs but nothing seems to match this case.

[amit-singh]

what is the config on your access switches defined as VTP transparent. Please check the trunk config on that also. Plesae paste the config from both the distribution switch and the access layer switch.

regards,

-amit singh

[wim_depauw]

Hi ,

the config on the access switches is identical to the configuration of the distrib switches.

I found a bug which describes my problem but this is for a cat6000 .

CSCin69677

Could it be the same ?

[amit-singh]

No its not the same..

Please see this :

Error Message SPANTREE-2-RECV_1Q_NON_1QTRUNK: Received 802.1Q BPDU on non 802.1Q trunk [char] [char].

Explanation A Shared Spanning Tree Protocol (SSTP) bridge protocol data unit (BPDU) was received on the listed interface. The interface was in trunk mode but was not using 802.1Q encapsulation. is The interface ID is char.

Recommended Action Verify that the configuration and operational state of the listed interface and that of the interface to which it is connected are in the same mode (access or trunk). If the mode is trunk, verify that both interfaces have the same encapsulation (ISL or 802.1Q). Once these parameters are consistent, spanning tree automatically unblocks the interface as appropriate.

http://www.cisco.com/en/US/products/hw/switches/ps663/products_system_message_guide_chapter09186a00800dea21.html#1186226

again please paste the config of the switches.

regards,

-amit singh

[wim_depauw]

Hi Amit,

I saw this message also and it does work fine with distribution switches with only a couple of access switches .

I can't put the total config of switches on the web because of security reasons .I hope you understand this.

We are running rapid-pvst on the distribution switches and pvst on acces switches .In the near future I'll change this to rapid-pvst.

spanning-tree mode rapid-pvst

spanning-tree extend system-id

spanning-tree pathcost method long

spanning-tree vlan 20 priority 24576

gr

wim

[amit-singh]

Hi,

Use the command " show interface fast x/y switchport " on the interface of both the access and distribution layer switches and see if the status is trunking and also matches.

Also, this could be due to the fact that you are runing rapid-pvst on distri. and pvst on access layer switches. If you ahve RSTP configured make sure that it is enable on all the switches in the domain ortherwise you will see unexpected results.

HTH,

-amit singh

[wbreitenmoser]

hello. why you have set the command "switchport access vlan 20 " on this interface? i think that could be the reason....