STP Designated Port Question

huafist

Hi everybody,
I'm currently studying to take the 640-802 exam, and I'm using the ExamCram 3rd Edition book to prepare for it. So far, I've noticed some pretty glaring mistakes in the book, but I have been able to verify with math or logic that there are errors. However, now I've ran into a problem - I don't know if there's a mistake, or if I'm just not getting something. Here is the diagram:



The question is:
"Which of the following will be the Designated Port for the ethernet segment between switches C and D?
A) Switch C, Fa0/1
Switch C, Fa0/2
C) Switch D, Fa0/1
D) Switch D, Fa0/2

The answer provided is:
Answer D is correct. Because D has the lower Root Path Cost (at 19) than C (at 23), D will make its port the Designated port - even though C has a lower Bridge ID.

Now, my question is: If there is a direct 10Mbit link between C and D (at a cost of 100), then with the cost the same in both directions, wouldn't MAC value determine which switch had the Designated Port for that segment? It looks to me like the costs that they show in the answer is the cost from the link through the root switch. Am I missing something?

Thanks for the help guys.
--
Rob Hall

drkat

Switch B's Bridge Priority is 1024 - This will be the Root Bridge

Cost from Switch B Gi0/12 to Switch D is 42

Switch B Fa0/1 to Switch C cost is 38 so this is a lower cost to go through Switch D to get to Switch C which is why Switch D has the designated port

(the hub is just an extension of the same link)


During STP election the priority's are compared first and only if there is a tie does the mac address come into play. With all costs being equal and bp's being equal yes the mac's come into play.

Forsaken_GA

drkat is right.

The other way to look at it is like this -

The link between C and D through the hub is the point where the topology loops. Take that link away, no loop. Add that link, loop.

That tells you that one of those two ports is going to be in Blocking mode. Who blocks? The port with the worst path to the root, which is Switch C Fa0/2, mostly because Switch D has the advantage of being directly connected to the root, while Switch C has to transit another one.

huafist

Gotcha. I understand where my logic was wrong now. Thanks guys.

Forsaken_GA

Now, to take it further... what I said about that particular link being the looping point in the network is a matter of perspective.

You know the root bridge, so you know all of switch B's ports will be designated. From there, you can fill in some of the root ports (ie, anything directly connected to the root).

So the question then becomes switch C, since it's the only switch that's not connected to the root. It's root port will be whichever port is the lowest cost to the root. Given the costs on the chart, it's root port will be the port connected up to Switch A. If you fiddled the costs so that it preferred the connection via Switch D, then that port becomes a Root Port and the other one transitions to blocking. But even in that case, Fa0/2 on switch D will still be a designated port, so in this scenario, the only thing that Fa0/2 on Switch D *can* be is Designated, unless you raise the cost on it's root port to an absurdly high value so that Switch D would prefer to transit 2 other switches instead of go directly to the root, and that's bad design.

huafist

Forsaken_GA wrote: »

Now, to take it further... what I said about that particular link being the looping point in the network is a matter of perspective.

You know the root bridge, so you know all of switch B's ports will be designated. From there, you can fill in some of the root ports (ie, anything directly connected to the root).

So the question then becomes switch C, since it's the only switch that's not connected to the root. It's root port will be whichever port is the lowest cost to the root. Given the costs on the chart, it's root port will be the port connected up to Switch A. If you fiddled the costs so that it preferred the connection via Switch D, then that port becomes a Root Port and the other one transitions to blocking. But even in that case, Fa0/2 on switch D will still be a designated port, so in this scenario, the only thing that Fa0/2 on Switch D *can* be is Designated, unless you raise the cost on it's root port to an absurdly high value so that Switch D would prefer to transit 2 other switches instead of go directly to the root, and that's bad design.

Very good information. Thank you. I'm pretty sure I'm solid on that now ; but I have another question that I'm going to start a new thread on.