stp, root port does not receive bpdu for 20 sec

[sarahr202]

Hi every body!

I have this question.

Let say three switches, sw1,sw2 and sw3 are connected in traingle, with switch sw1 as a root bridge.

Sw2 for some reason, stops receiving bpdu on its root port for more than 20 seconds.

1) what would happen to root port now? will it become designated port?

2) will the root port though not receiving any bpdu, will remain in forwarding state?

thanks a lot!

[Diego Vargas · Accepted solution]

Hi,

1) what would happen to root port now? will it become designated port?

Yes

2) will the root port though not receiving any bpdu, will remain in forwarding state?

Yes

Since sw2 stops receiving BPDUs from sw1 it would assume that he is the Root but sw3 will advise that sw1 has better priority and will tell sw2 that sw1 is root and can be reach though him.

The port in Sw2 connected to sw3 will become the root port and will be forwarding while the port to sw1 will be designated and sending BPDUs. This will cause a loop

The way to prevent this issue would be to use UDLD. This feature would prevent the problem since the issue is not that the link is down but that not BPDUs are received, this could be caused by Unidirectional problems (either caused by a software issue or a problem on the fiber), then UDLD should prevent a loop from showing up.

[sarahr202]

or we can use loop guard feature as well.

[asdvads vwe]

sarahr202, if the Root Port is not receiving BPDUs from a Designated Port on the other side or Root Bridge which is also for (Designated Port), then the Root Port will think by itself: "Hang on, we got a problem, I haven't heard from the DP for 20 seconds, he must be dead or there is a problem, I will transition my BLK Port into de Designated (Forwarding state). So this is what it really happens, but yes, we also can configure this Port with the Loop Guard feature which will force the port to stay in BLK mode even it haven't heard for 20 seconds from DP.

[sarahr202]

I think best option is to use loop guard. Here is why:

Udld may work or may not work.

If Udld is configured on sw1 and sw2 and sw2 is receiving udld messages but not bpdu from sw1, then port won't be put in erridisable state and the loop would form.

If sw2 does not receive udld message, asuming aggressive mode is configured then sw2 will send udld message every second for 7 seconds , if sw2 still does not receive any echo from sw1, then sw2 put its port(root port in our case) in erri disable state.

thanks alot!

[Mike Soha]

Can we expand this topic to discuss the scenario of the loss of the Root Bride, SW 1?    Thanks