STP Designated Ports

jscimeca715

Hello all, I've started reading Wendell Odom's ICND2 book and have a question about designated ports. Actually, it's more of a clarifying question, but a question nonetheless.

I think this diagram might be confusing, but I'm going to try anyway.

Switch A--Switch B--Switch C--Switch A (Imagine that this is a triangle and you should get the picture)

Switch A is the root switch, and Switch B and Switch C have designated their root port. Now, for the connection between Switch B and Switch C, Switch C has a lower cost so it's put into forwarding and the Switch B port is put into blocking.

So, what I'm trying to get at, is that Switch C's port on that segment is the Designated Port. But that doesn't mean the port is going to be used for anything unless the connection between SWA and SWC goes down correct?

I know this is probably confusing so please let me know if I'm completely on the wrong track or not! Thanks in advance.

/usr

But that doesn't mean the port is going to be used for anything unless the connection between SWA and SWC goes down correct

Unless I'm wrong and someone else corrects me...

You've got it right. The DP on Switch C is simply the one that has been elected to stay up, since both ports that create that segment do not have to be down.

That segment would not be used unless there is a change in the STP topology and Switch B brings up it's link.

tim100

jscimeca715 wrote: »

Hello all, I've started reading Wendell Odom's ICND2 book and have a question about designated ports. Actually, it's more of a clarifying question, but a question nonetheless.

I think this diagram might be confusing, but I'm going to try anyway.

Switch A--Switch B--Switch C--Switch A (Imagine that this is a triangle and you should get the picture)

Switch A is the root switch, and Switch B and Switch C have designated their root port. Now, for the connection between Switch B and Switch C, Switch C has a lower cost so it's put into forwarding and the Switch B port is put into blocking.

So, what I'm trying to get at, is that Switch C's port on that segment is the Designated Port. But that doesn't mean the port is going to be used for anything unless the connection between SWA and SWC goes down correct?

I know this is probably confusing so please let me know if I'm completely on the wrong track or not! Thanks in advance.

Designated ports are always in the forwarding state. A port that is blocked is an alternate path to the root. If you set the cost on a port to a lower value it becomes the root port. Root ports connect to designated ports. If switch A is the root switch its ports are designated ports. Switch B and Switch C's ports connecting to Switch A are root ports if you leave everything at the default. If you set the cost on Switch C's port connecting to Switch B to a lower value it will become the root port and it will block it's port connecting to Switch A. So the topology will be: Switch C --- Switch B --- Switch A (root).

EdTheLad

tim100 wrote: »

If you set the cost on Switch C's port connecting to Switch B to a lower value it will become the root port and it will block it's port connecting to Switch A. So the topology will be: Switch C --- Switch B --- Switch A (root).

Don't forget the cost is cummulatve, so the cost on switch B's port(B->A)would have to be lowered too, so that the cumulative cost from C->B->A is lower than the cost from C-> A.
Then the port to C via B has the lowest spt cost and will become the root port.

tech-airman

jscimeca715,

jscimeca715 wrote: »

Hello all, I've started reading Wendell Odom's ICND2 book and have a question about designated ports. Actually, it's more of a clarifying question, but a question nonetheless.

I think this diagram might be confusing, but I'm going to try anyway.

Switch A--Switch B--Switch C--Switch A (Imagine that this is a triangle and you should get the picture)

Switch A is the root switch, and Switch B and Switch C have designated their root port. Now, for the connection between Switch B and Switch C, Switch C has a lower cost so it's put into forwarding and the Switch B port is put into blocking.

I would be VERY careful using terms like "...designated their root port..." Switch A is the root bridge (to use STP terminology) and so has designated ports. The way I remember that is the root bridge is connected to the designated bridges by the designated ports. So now that the root bridge, in this case Switch A, has chosen it's designated ports, it's time for the designated bridges to decide amongst themselves who are the designated bridges (to use STP terminology). The designated bridges are the closest to the root bridge. In this case are Switch B and Switch C. The port on Switch B is connected to Switch A is called the root port because like above, the designated bridge port pointing towards the root bridge is the root port. Also, the port on Switch C that is connected to Switch A is also a root port. Then STP decides to block ports between designated bridges and any other non-root bridge interconnects, like the link between SWB and SWC.

jscimeca715 wrote: »

So, what I'm trying to get at, is that Switch C's port on that segment is the Designated Port. But that doesn't mean the port is going to be used for anything unless the connection between SWA and SWC goes down correct?

No, it is incorrect to state "...Switch C's port on that segment is the Designated Port." The root bridge has designated ports not designated bridges. In the case the link betwen SWA and SWC goes down, STP is run again to re-elect a new root bridge then designated bridges are appointed by the new root bridge. If the link between SWA and SWC goes down, then yes, the link between SWB and SWC becomes important.

jscimeca715 wrote: »

I know this is probably confusing so please let me know if I'm completely on the wrong track or not! Thanks in advance.

STP can be confusing when the terminology and STP process are not properly understood.

jscimeca715

Thanks everyone, I studied it for about an hour last night and it started to make sense. Your examples just confirmed my study, I appreciate the help!

aordal

Theres a couple 45 minutes videos on Cisco Systems, Inc that explains both STP and RSTP pretty well. You have to sign up, but its' free. I recommend taking some time and watching them.

drew2000

aordal,
After doing some searches on Cisco.com and the Cisco learning center, I couldn't find any links for STP Videos.

Do you happen to have the link or do you have the title of the video so we can search again?

thanks,
Drew

Neeko

I know this is an old thread but I've just been looking for some info on STP and came across this thread, and wanted to clarify something.

tech-airman wrote: »

jscimeca715,
No, it is incorrect to state "...Switch C's port on that segment is the Designated Port." The root bridge has designated ports not designated bridges. In the case the link betwen SWA and SWC goes down, STP is run again to re-elect a new root bridge then designated bridges are appointed by the new root bridge. If the link between SWA and SWC goes down, then yes, the link between SWB and SWC becomes important.

From my understanding it is perfectly acceptable to refer to a port on a non-root switch as a 'designated port' since each segment will have one. It isn't just root bridge that has designated ports and a redundant link between two switches is not necessarily blocked by both, in the case of a single trunk one switch will block and one will forward (the designated port for that segment).

Anyway... on a side note when the original 802.1d protocol is running, in the A-B-C-A triangle topology, when the link between B and C is redundant and B's interface on that link is the designated port, what exactly happens if the link between A and B goes down? Will B continue to send BPDUs to C, as I thought these were only sent when received from the root bridge which B no longer would, meaning C will realise it no longer has the higher cost to reach the root on that segment and will therefore assign its interface as the designated port and switch B will receive BPDUs from C and reconverge?

Or does B stop sending BPDUs as it is not receiving them from the root, and when C notices that it is no longer receiving BPDUs it will set its interface to forwarding as it can assume it has the lower cost path to the root, allowing B to learn the lower cost and reconverge?

Basically, In the scenario where B's interface that connects to C is blocking, it will still receive the BPDU from C and be able to reconverge. But when it is C's interface that is blocking, how does C decide to unblock that port to allow B to reconverge?

Hopefully it makes sense what I'm asking. I was going to try and work it out for myself in packet tracer using the debug spanning-tree events command but it doesn't seem to be supported.

kryolla

B will start to announce itself as the root because it lost its connection to the root bridge, C will hear and ignore an inferior BPDU on its blocking port, it will max age the stored BPDU and C will transition that port to forwarding or designated. You can probably find an article that will get more detailed but that the jist of it.

edit look into backbone fast to speed up this process

I just found this link

http://www.cisco.com/en/US/tech/tk389/tk621/technologies_tech_note09186a00800c2548.shtml

Neeko

kryolla wrote: »

B will start to announce itself as the root because it lost its connection to the root bridge, C will hear and ignore an inferior BPDU on its blocking port, it will max age the stored BPDU and C will transition that port to forwarding or designated. You can probably find an article that will get more detailed but that the jist of it.

edit look into backbone fast to speed up this process

I just found this link

Understanding and Configuring Backbone Fast on Catalyst Switches - Cisco Systems

Excellent, thanks. That link uses the exact same topology, DP and point of failure I was using.

phoeneous

tech-airman wrote: »

jscimeca715,



The port on Switch B is connected to Switch A is called the root port because like above, the designated bridge port pointing towards the root bridge is the root port. Also, the port on Switch C that is connected to Switch A is also a root port.

The root port of a designated switch isnt decided simply because it points to the root bridge, it is decided based on the total cumulative cost to the root bridge.

So if the cost from Switch B (fa0/1) to Switch C to Switch A is 12 and the cost from Switch B (fa0/2) to Switch A is 19 then Switch B's fa0/1 interface will be the root port.

boggle

/usr wrote: »

Unless I'm wrong and someone else corrects me...

You've got it right. The DP on Switch C is simply the one that has been elected to stay up, since both ports that create that segment do not have to be down.

That segment would not be used unless there is a change in the STP topology and Switch B brings up it's link.

Just to clarify further - the B-C segment may well be used if there is other stuff on it.

Although today most segments are just point-to-point ethernet links (so if if the segment joins two switches and one of the ports is blocking, nothing happens on the link), historically they would be LAN segments with possibly many devices sharing access to the segment. Imagine a few years ago when switches were expensive and multi-access was through hubs or real live mutli-drop thinwire ethernet. There could be a dozen PCs and three switches connected to that LAN segment.

The 'designated port' would be the only switch port on the segment that would accept frames from anything on the segment and would forward them out of its root port and any other designated ports (unless it was a unicast frame and the switch had learned exactly where to send it).

Broadcast or unlearned unicast frames arriving at a non-blocking port (root or designated) are forwarded out of all other non-blocking ports. That way they reach all switches and all segments and hence all devices on the LAN. The STP blocks enough ports to ensure this broadcast doesn't loop and cause a broadcast storm.

dmcneil330

I feel like some of these answers are a bit confusing.

I think no one answered this question:

So, what I'm trying to get at, is that Switch C's port on that segment is the Designated Port. But that doesn't mean the port is going to be used for anything unless the connection between SWA and SWC goes down correct?

It isn't accurate to say that these ports won't be used for anything. Designated ports are also used for sending BPDUs to downstream switches. So switch C's port will send BPDUs to switch B's blocking port at regular intervals. Switch B's blocking port will receive those BPDUs and expect to receive them. If it stops receiving them then it will hold on to the last BPDU for Max-Age time and then attempt to transition.