View Full Version : Another STP question...
echoreply
07-02-2008, 01:15 PM
Now studing for the BCMSN I have to refer to the Ciscopress books :mad:
My question refers to a scenario where PC1 is connected to 2 switches for redundancy and forwarding a frame to the PC4 on the other side.
I understand the concept of a switching loop ...but this has me hung up.
It says "lets assume that both switches know the mac address both of the PCs in question. It then states that "both switches will receive the frame on their incoming ports". Then it will forward the frames on to PC4 and PC4 will receive both frames. It states this is not ideal, but not disasterous either".
Question 1 : How/Where did the frame get duplicated? (the frame is received on both incoming switch interfaces)
Am I missing somthing here?
Any help is appreciated.
pconlan
07-02-2008, 01:19 PM
What version of the book are you using, and what is the page number. I don't use those books very much and I will have to see if I have it or not. If I don't, I may need a little more information about the problem.
Thanks
Pat
echoreply
07-02-2008, 01:23 PM
The CiscoPress CCNP BCMSN Official Exam Cert. guide. (I bought the whole library)
Fourth Ed. By David Hucaby. Page 187 Redundant Brindging w/ 2 switches.
Thanks for your prompt help... this really has me at a stopping point because the next 3 chapters are STP in detail.
crissa
07-03-2008, 03:44 AM
The CiscoPress CCNP BCMSN Official Exam Cert. guide. (I bought the whole library)
Fourth Ed. By David Hucaby. Page 187 Redundant Brindging w/ 2 switches.
Thanks for your prompt help... this really has me at a stopping point because the next 3 chapters are STP in detail.
Hello,
my book was printed in June 2007.
As I understand you a referring to Figure 8-2 and the text direct under the figure.
The frame is duplicated because both switches received it and so both switches will send it out! As both switches know the mac address of PC-4 they can send out the frame without further asking.
BTW, maybe it helps (you) to read the STP in Todds book again. ;-)
Bye, Tore
echoreply
07-03-2008, 07:11 AM
I understand that both switches receive the frame... How is that when only 1 frame is transmitted.
That is the standard book answer. From what I understand a switch adds the MAC and Dl back to the bits from the physical. Nowhere does the book say that if two switches are on the network the frame is duplicated. If it is when and how?
crissa
07-03-2008, 10:41 AM
I understand that both switches receive the frame... How is that when only 1 frame is transmitted.
That is the standard book answer. From what I understand a switch adds the MAC and Dl back to the bits from the physical. Nowhere does the book say that if two switches are on the network the frame is duplicated. If it is when and how?
Hello,
because Segment A (as Segment B) allows the transport of this one frame to both switches at the same time. If you think of the segment as of a BNC cable or a hub this can work.
HTH!
Bye, Tore
echoreply
07-03-2008, 10:50 AM
So.. If I understand this correctly... (and I am reconsulting Lammle's book) the frame when placed on the wire is visible to all layer2 devices. Nothing is duplicated it just is visible and logged into the MAC table at that point.
lammle
07-03-2008, 11:03 AM
Every device connected into a switch port will see the frame when the switch send it out. The switch will only send it out one port, unless it is a broadcast or mutlicast frame.
If you have a hub with 10 devices connected into a switch port, all 10 devices will see every frame sent out the switch port.
Todd
crissa
07-03-2008, 11:16 AM
So.. If I understand this correctly... (and I am reconsulting Lammle's book) the frame when placed on the wire is visible to all layer2 devices. Nothing is duplicated it just is visible and logged into the MAC table at that point.
Hello,
the text states that both switches know PC-1 and PC-4. So both switches know where to deliver the frame.
Maybe you think to technical, this is just an example (and maybe not a good one).
Bye, Tore
echoreply
07-03-2008, 11:17 AM
Please bear w/ me for just a minute. Let me make sure I understand this.
My issue is the CP book says that when PC1 sends a frame to PC4 and both switches know the topology, the frame from PC1 is rcvd on both interfaces of both switches.
At the same time? That seems somewhere the frame was duplicated. I would think the first switch that rcvd the frame would process it if it is in its MAC table.
I am just not getting it i guess. I tend to obsess over details.
echoreply
07-03-2008, 11:21 AM
I appreciate eveyones patience helping me understand this concept.
lammle
07-03-2008, 11:37 AM
ER, I don't have that book so I can't respond to that exact issue.
However, the only way that a frame from a host will go to more then one switch is if there is a duplicate connection, which is not right.
I guess I need to see the figure before I can understand the problem here.
I think I have that book in my Dallas office, which I'll be in next week.
Todd
lammle
07-07-2008, 03:31 PM
Okay, I got the book.
He is explaining basically what not to do.
You don't want a redundant design like this. It will create switching loops, duplicate frame copies and broadcast storms.
If you left the defaults on, STP would stop this from being a problem.
He is explaining basically what would happen if STP was disabled and you had redundant links. Bad idea.
Todd
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.