Search this Blog

Tuesday, January 18, 2011

What are the different spanning-tree terminologies and how are they different from others?

We have some confusion of the spanning tree terminologies We came across while studying. Can someone explain clearly on what they and how different they are from others?

1. PVST – We know that this is the legacy STP protocol (cisco proprietary) and will run on ISL trunk links. But in my router we see only 3 modes for STP namely PVST, MST and RAPID-PVST. So if we choose to run pvst mode over dot1q trunk links , do we essentially run PVST+ or still we run only PVST?

2. PVST+ - Continuation to above question how to run PVST+ mode on our routers we see only the above mentioned 3 modes available.

3. Spanning tree enabled protocol ieee - When we saw this when we execute "show spanning-tree vlan x" which mode do we assume PVST or PVST+?

4. CST - What is a common spanning tree and why it is needed? read in some document mentioning that is will run on STP instance for all VLANs to reduce switch load but does it essentially mean a flavor of MST mode?

5. IST - Why do we need an IST instance inside an MST regions? It says that it will only generate the BPDUs inside a region but how will it differentiate different instances inside MST? Any useful document link or video link here would be very useful to me.

6. rapid-pvst - When we choose to run this mode do we essentially run r-pvst or r-pvst+ mode?

7. Root bridge selection – We know the process of a root bridge selection but my doubt is why when a switch with a lowest bridge is added to the network again an election happens? We are comparing this to DR/BDR election in OSPF network where the re-election will happen only when the already elected routers fail.

8. BPDUGUARD Vs BPDUFILTER - What is the difference between these two features? Both of them are anyways filtering the bpdu then why two separate features?

9. Aging Time 300 – We see this under "Bridge ID" section of "show spanning-tree vlan x" output. What does this signify and how this is helpful? The mac-address table aging time?

10. show spanning-tree bridge id - The output of this command is as follows.

nPE1#show spanning-tree bridge id
VLAN0001 8001.0021.d8ca.c340
VLAN0002 8002.0021.d8ca.c340

What does the 8001 and 8002 signify? We were thinking of bridge priority (with vlan number) and mac-address as BID of a switch. But this command yields some different output.

11. RLQ request/response – We know that these messages are used in backbonefast feature but are they a different type of BPDUs or it is embedded in conf BPDUs? To my knowledge only 2 types of BPDUs exists. Configuration and TCN BPDUs. Are these are different kind? WEtried to captures after enabling the backbonefast feature but we could not find in my wireshark. We turned on the debugs where it says that it received rlq messages but we couldnt see in packet capture

12. STP/R-STP Timers - In legacy stp, all switches send the BPDUs to the root and the root inturn propagates to all other switches. So why the timer configured on root bridge will take effect on all swithces. But in RSTP all switches all the capability to send BPDUs then if there is a mismatch in STP timers between itself and the root, which one will be agreed upon?

Tips:

1. If you are running spanning-tree mode pvst then the PVST or PVST+ will be automatically run on trunks according to their encapsulation. ISL trunks will run PVST while IEEE 802.1Q trunks will run PVST+. There is no special command to activate only PVST or only PVST+.

2. Simply make all trunks do 802.1Q encapsulation and if the spanning-tree mode pvst is configured, the PVST+ will be running automatically on the trunks.

3. Assume both, as the exact type is decided on a per-trunk basis depending on its encapsulation.

4. Originally, the 802.1D and 802.1w STP/RSTP as specified by IEEE were not VLAN-aware. What they created when run in a network was a spanning-tree whose topology was identically shared by all VLANs, hence the term 'Common Spanning Tree' - common for all VLANs .Agreed, it is in effect quite similar to running MST using a single instance. Cisco switches use the CST when talking to non-PVST world.Without going into lengthy explanations, let me reference a document that may clarify this better than I can at the moment: click here for the document.

5. By far the best explanation of the entire MST was given by Petr Lapukhov on his INE blog - enjoy here: Click here for blog 1 . Click here for blog 2. While not an easy reading (the MSTP is not an easy topic!), I suggest very much reading these two articles first, and then please come back with any additional questions about the MSTP. These articles will very certainly answer lots of your questions here.

6. Both on a per-trunk basis, depending on the encapsulation used.

7. Indeed, when a new switch is added to a network whose BID is lower than that of the current root, new elections will take place and the added switch will become the new root switch. If the added switch has a higher BID, no election occurs. Your question has a very good merit - is it actually appropriate for the STP and all its incarnations to preemptively change the root switch? That is a question that will probably go unanswered as there are good arguments both for and against a preemptive election. The DR/BDR elections are non-preemptive in OSPF, and the DIS election is preemptive in IS-IS (both DR and DIS have very similar functions), so this is a nice example that it can be done both ways. For now, I can only say that the STP was designed that way, and no one challenged that so far.

8. They are quite dissimilar, actually. The BPDUGuard puts a port into an err-disabled state when a BPDU is received. It is usually put on edge ports as an additional protection technique against possible loops or unauthorized extensions of a network. Should a BPDU be received (no normal workstation ever sends BPDUs!), it is clear that there is something inappropriate connected to that port, and it will be shutdown.The BPDUFilter prevents BPDUs from being sent and received on an interface. The BPDUFilter is essentially a way how to "turn off" STP on a particular port. A port with BPDUFilter activated directly in the interface configuration won't send any BPDUs, and will ignore all received BPDUs as if they never came. The BPDUFilter can be also activated globally using the command spanning-tree portfast bpdufilter default in which case it applies only to PortFast-enabled ports, and it behaves differently in that case: when a port comes up, it sends 10-11 BPDUs and then it stops sending BPDUs. If it ever receives a BPDU, the BPDUFilter (and the PortFast as well) is deactivated on that port and the port becomes a normal non-edge STP-participating port that receives and sends BPDUs just like any other port - until the port is disconnected. While these two features may seem similar, they are quite unlike. As a single example, a BPDUFilter by itself will never cause a port to become blocking or shutdown while the BPDUGuard will make just that.

9. Basically, this output tells whether the aging time for MAC entries of a particular VLAN are aging according to the global configuration of the switch, or whether the switch is currently receiving the TC BPDUs (Topology Change), in which case the MAC address aging time is reduced to the forward_delay value advertised by the current root switch.

10. The 8001 and 8002 is the bridge priority. With all recent switches, the bridge priority has a configurable value (in steps of 4096) and a fixed value into which the VLAN number is copied (this functionality is called the extended system ID). The 0x8001 = 32769 and 0x8002 = 32770, signifying that the priority of the switch for both VLANs is 32768 (the default value) and the VLAN IDs are 1 (32768+1) and 2 (32768+2).

11. Click here for more information:

12. If I am not mistaken, even in RSTP, the frequency of sending BPDUs is dictated by the timers of the current root switch.

Citation - This blog post does not reflect original content from the author. Rather it summarizes content that are relevant to the topic from different sources in the web. The sources might include any online discussion boards, forums, websites and others.

No comments :

Post a Comment

 
/* Google Analytics begin ----------------------------------------------- */ /* Google Analytics end ----------------------------------------------- */