A way to pass BPDU's?

[Uintadave]

Is there a way to allow the switch to ignore BPDU's and just let them pass? I can not find a way to have the switch do this. In order to patch this shortcoming I had to place a switch in front of the switch on the trunk ports so that BPDU's could be seen on the other trunk ports.

Here is a basic setup.

We assign a VLAN per port so that the trunk can move around what port is used by simply moving the VLAN on the interface. This has been very useful when say a SFP interface decides to act up and a good way to control pause frames. What we need is BPDU's to be un altered and allowed to pass from one trunk port to another. With it blocking the BPDU's it causes RSTP to have an incorrect topology and results in loops. When I put a dumb switch on the trunk ports before plugging them into the netonix the BPDU's were passed and it worked correctly.

So is there a CLI or some way of telling the switch to not mess with BPDU's? We are using FW 1.5.5

[Stephen]

viewtopic.php?f=17&t=5687&hilit=BDPU&start=10

[Uintadave]

Stephen, Thanks for the reply. I still need a little bit of clarification.

Is it possible to enable BPDU passthrough on an Netonix switch? Or is this a feature that you are refusing to allow because you do not see a use for it? It appears others have needed it and I do need it now.

[mike99]

No, the switch don't support L2CP that allow to encapsulate L2 protocols.

[Uintadave]

Thank you Mike for the answer.

I suppose I have two options for this setup then.

I can put a switch in front of each switch to get the desired functionality or

I can just use something like a CRS328-24P-4S+RM for the AP's. Sure it can't power a AF24 or a Siklu but I will only need the RSTP for the access not links, and it has SFP+.

Thank you for the quick response.

[mike99]

Or use a passive PoE injector. WS3 have a carrier class switch chip that support more features (not sure which one exactly since I don't have the vitesse doc). Maybe the new line could support it eventually.

[LRL]

I would not think passing BPDU's when the switch is not STP aware as a mistake merely the proper function as anything that is actually stp aware would only detects loops if BPDUs are passed through.

In much older Netonix firmware versions BPDU's were passed through if STP was disabled. We are looking for this functionality that existed years ago. If memory serves the switch chip would yank BPDU's once it had STP config pushed to it, if there a way to remove that achieve this?