01-07-2008 05:21 AM - edited 03-05-2019 08:19 PM
Hi,
my question is in order to clarify the standard STP functionality.
Scenario:
three switches connected in a
ring and so forming a loop;
only one of the six ports linking the switches has STP enabled (other ports have STP disabled).
Question: is the switch able to detect the loop and so blocking the STP enabled port ?
If no ...in this scenario how many of the ports must be STP enabled ?
Thanks in advance.
Pierluigi
Solved! Go to Solution.
01-07-2008 10:37 AM
I agree, it all comes down to what you mean by disabling STP. If the switches that have STP disabled are flooding BPDUs like user traffic (dumb switches that act like hubs do that, PVST+ switches with "no spanning-tree vlan X" configured attempt to do that also), then the unique bridge running STP should be able to see its own BPDUs coming back at itself around the ring and prevent the loop. Note that you could still have temporary loops when a failed link is recovered, and that convergence will be slow (based on timers).
Regards,
Francois
01-07-2008 05:39 AM
Hi,
Only ports partecipating in spanning tree move between different states:blocking, listening, learning and forwarding.
You have to enable spanning tree on 6 uplink ports, in order to detect loop, and then have a topology loop-free.
For more information look at: http://www.cisco.com/en/US/tech/tk389/tk621/technologies_configuration_example09186a008009467c.shtml
and
I hope this helps.
Best regards.
Massimiliano.
01-07-2008 05:51 AM
Yes STP is active & thats the reason why some ports are blocked. However, I'm not sure of the statement, only 1 port is active out of six ports. You may check this by "sh spanning-tree".
01-07-2008 05:58 AM
If you have disactivated STP on the ports by using bdpufilter, then you will have a loop. The BPDU will not make it round the loop, so the loop will not be detected.
Kevin Dorrell
Luxembourg
01-07-2008 10:37 AM
I agree, it all comes down to what you mean by disabling STP. If the switches that have STP disabled are flooding BPDUs like user traffic (dumb switches that act like hubs do that, PVST+ switches with "no spanning-tree vlan X" configured attempt to do that also), then the unique bridge running STP should be able to see its own BPDUs coming back at itself around the ring and prevent the loop. Note that you could still have temporary loops when a failed link is recovered, and that convergence will be slow (based on timers).
Regards,
Francois
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide