vlan BDPU issues

moved my network from L3 to L2.

Saw an issue please help:

(CUST 1)
||
(SM Vlan 994)
||
( CANOPY CLOUD1 ) (CANOPY CLOUD 2)
|| || || (all connections to 3550 are trunk encap 802.1q
(CAT1 3550) (CAT2 -------3550)

CAT2 is the VTP server.
Our customer has 3 sites connected to the same cluster. On their main site they are terminating into a 2950 switch, with complete default settings. They have been working fine for 2 months, we switch to L2 and they go down.

The port is blinking orange and console output complaining about BPDU inconsistent naive VLAN.

I thought my SM would have stripped all tagging before sending data to the customer switch,but somethign is affecting it. I went down and cahnaged the port from switch to trunk and all of a sudden it started working but learned all our vlan information. Also it somehow joined our vlan domain and started acting as a server, so we have 2 servers and thus brought down the network.

What do I need to do to get this cisco switch to operte as a non-manged switchport and secondaly how can i stop a customer sticking in a switch and bringing down my network. I have all all frames on the SM running 7.2.9.

I am new to vlan stuff.

on 2950, on FE port to SM set
switcport mode access
switchport access vlan x, where X is the customer VLAN that traffic from SM should go into.

On SM set
Untagged Ingress VID to VLAN assigned to customer

hi mkecman,

this is what happened, the 2950 was on complete default setting, customer was configured on vlan 994, the port went orange.

After much playing about the only way I got it to work was take out vlan 994 from my main VTP server switch.

Would the management ID have anything to do here.