144 IPExpert Rapid And Rapid Per VLAN Spanning Tree Converge

[Music] the thing that truly makes the rapid spanning tree rapid is its convergence process in other words how does the rapid spanning tree calculate the resulting tree like structure in a fast and efficient way and what makes it faster than the traditional spanning tree the rapid spanning tree convergence consists of synchronization process which in turn consists of an exchange of messages that we call proposals and agreements a very common misconception that students have when they are learning about the rapid spanning tree is that rapid spanning tree is faster or more rapid than the traditional spanning tree because of faster timers this is not the case as the timers in used by the rapid spanning tree are pretty much the same timers that are used by traditional spanning tree what truly makes rapid spanning tree rapid are these proposals and agreements so let's see how they work the process of rapid pv st convergence is done using this exchange of proposals and agreements this process is technically consisting of two steps one is the handshake the agreement and propel proposal and agreement exchange between two switches in the other one is the process of synchronization the process of synchronization is what the switch does before it agrees to anything that others which may have proposed but I keep mentioning these proposals and agreements without actually talking about them in greater detail bear with me the first thing that I want to mention here before I go on is that this process works only on point-to-point ports by default those ports that are not point-to-point at our shared ports may need to be configured manually to be point-to-point ports and remember this is going to be the case on all ports that are not full duplex and they can be manually configured to be point-to-point ports the reason why this works on point-to-point port is that this whole idea of handshake of the exchange of these proposal and agreement messages is supposed to be happening between only two devices if we have more than two devices involved on the same network segment we cannot use this process and if we have half duplex connection that may imply that we have a hub and we if we have a hub that means that we have multiple devices connected to that Network segment and of course we can use this process so this is the first safeguard that is put in place the next thing that happens is if this has been satisfied so if we have this condition in place is that the handshake happens and it's a relatively simple process the switches MVPD use and they propose that this port that has actually transmitted the BPU becomes a designated port the other switch may actually send a superior BPU a better baby new or it can agree to the proposal now let's say that the other switch does agree to the proposal before it blindly sends out the agreement and says yeah this is okay you'll be the designated port what it needs to do is it means to synchronize itself with the topology because remember we may be dealing with more than just two devices if we are dealing with more than just two devices blindly trance changing the mode of operation on the port transiting from down to forwarding or from discarding to Forli could be dangerous it could lead to bridging loops and we want to avoid the bridging groups so before the switch can send the agreement message it actually needs to synchronize itself with the topology what does that mean what this means of course is that the switch needs to go through the process this process of these it consists of several steps the first thing that happens is that all non edge ports are moved into the discarding State the second thing that happens on the port on which we have received the proposal to which we are just agreeing to we are actually going to send out the agreement so we are going to say yeah okay you are going to be the designated port on this segment and I'm now synchronizing myself so all the non edge port are at this moment discarding the traffic now the next thing that we are going to do is we're going to determine which one of our ports is going to become the root port the root port remember is the port that is closest to the root that has the lowest cost to reach the root this port is immediately moved to forwarding because we don't want to actually cut ourselves off from the network for too long period of a time now on each non edge port that is currently in this cutting State as the result of this process of synchronization we are going to send our own proposals so basically imagine what happened here the switch sent out the proposal this switch here is now agreeing to this proposal and it's sending its own proposals out let me show you this on the whiteboard so what we have here is we had the first switch and our second switch here so this was the link on which we initially received the proposal message now we are agreeing on this port here but imagine that we had other ports here that other switches may have been connected to so on these ports here the first thing that we did is removed all these ports into discarding State and now at this moment we are sending our own proposals basically what we are doing at this moment on these which is here we are triggering their own synchronization process so if these switches have downstream switches here they're going to repeat the same process that we have started here so what happened here is we had the proposal we are sending the proposal and the proposal goes out just like this so once we send our own proposals we are basically going to be expecting and waiting for the agreements to arrive when receive the agreement message on the port on which we have sent out the proposal on we are immediately going to transition it into the forwarding State so basically what happens here in our example is when this switch here may have a we'd with the agreement at that moment we are going to start forwarding here so this port is no longer going to be blocking this one is blocking or discarding and this one is discarding but this port here has transitioned to become a designated port in a forwarding state so once that has happened there we have moved our port into the forwarding state and the synchronization on this switch has been completed let's now see all this in action to do that I'm obviously going to use some equipment and let me quickly show you what this equipment is going to be so to start with I'm going to configure two switches I'll be adding one more switch down the road but I can start with just two switches these are going to be my switch number one I'm going to call it cat one and I'm going to have cat to here they're going to be interconnected on fasting at 24 port and I'm going to configure these tools which is to run rapid prevail on spanning tree plus and I'm going to configure cat one to be the route and I'm going to make sure that cut one is the route by setting the lowest possible priority there I'm going to configure prayer it is zero on this switch also what I'm going to do here now this is not necessary but I really really want to simplify things for me I'm going to configure both of these ports here as access ports I don't have to do that I can leave them to be trunks but what I really really want is to see the behavior in just a single one now if I have multiple VLANs on the port this same process will happen multiple VLANs I'm just going to be seeing more bpdus than I actually need to examine the behavior that I'm after and the behavior that I'm after is actually the behavior on this port and in general on cat to switch I want to see what cat 2 does with its port how it transitions from one state to another and I want to see my handshake and synchronization process in action to do that I'm going to be you the combination of two tools I'm going to be using some debug commands on cat too and also what I'm going to be using is going to be my handy Wireshark which is going to be connected to Fast Ethernet 0 3 port here on cat 2 m to start with I'm just going to mirror the traffic from this port to my Wireshark station so let's go ahead and configure this before I show you the configuration I promised I'm going to show you something else something that I have not yet talked about what is going to happen if I configure one switch to operate in rapid mode and the other switch to operate in a non rapid mode well the answer to that question is very simple the switch configured to operate in rapid mode is going to fall back and operate in a traditional spanning tree mode luckily this is very very easy to test and prove I'm simply going to leave cat two in its default configuration it's going to operate in a traditional spanning tree mode I'm going to configure cat one on the other hand to operate as I described I would I'm going to make sure it's the root for villain one and I'm going to change it to rapid prevalence spanning tree mode let's go ahead and see what's going to happen now keep in mind that there are some timers involved and may need to move a little bit fast so what I'm going to do on cat one I'm going to change spanning tree node rapid previously I'm going to say spanning tree VLAN one priority zero so this is going to make sure that ket 1 is operating in rapid mode and that for VLAN 1 it is going to become the root well if ket 2 is in default configuration which it is then I'm going to go to interface fast init 24 and I'm going to say no shutdown very quickly here I'm going to say show spanning tree and I'm going to do the same thing on cat too now what I should be seeing here is all cat - I see that port moved into the listening phase but I can see that only cat 1 because there is no listening phase in rapid spanning tree this port is actually in block instead for the forward delay time after that expires it is actually going to move into the learning phase in the learning phase we are basically learning what are the MAC addresses on the other side and this is the same thing that happened on cat to get to normally transition to learning phase well as I said timers are involved I was a little bit too late so there was a learning state involved on cat 2 as well but now we are already in forwarding state which should be the same case on cat 1 so now you can see that this is a designated forwarding port but the important thing is that on cat 1 on the side that was rapid we had a transition from blocking state into the learning state and finally into the forward instead so we could not benefit in this case from our handshake and proposals and agreements because cat 2 was unable to answer any of this stuff because it was running the traditional spanning tree now let's configure this properly let's configure both sides to run the rapid spanning tree basically what we have to do is just change the configuration on our cat - so I'm going to go to cat - when I'm going to say spanning tree node rapid spanning tree then I'm just going to reconfigure my ports to be access ports so switch port node access they will be in VLAN 1 by default but let's make sure that I'll do this same configuration on both sides so spanning tree mode access here actually not spanning tree but switch port mode access let's bounce the port actually I'm just going to shut it down on both sides because I don't want to focus on what is happening on cat - as I said what I want to do here is I want to observe the debug output on cat - and I want to capture some traffic in my Wireshark connected to fastener 0 3 on cat - so I need to configure the debug and I need to configure my span session to make this happen so I'm going to go to catch - and first thing I'm going to do is make sure that fast in 8:03 is up now this is the port facing my Wireshark then I'm going to configure my spam section I'm going to say monitor session one source interface fasting of 24 and monitor session one destination interface fast image 0 3 and capsulation replicate because I want to preserve all the tags that might be going on there and I'm going to go to interface fastener 24 and I'm going to again make sure it is shut down then I'm going to go to my Wireshark oops let's go here and I'm going to start the capture on the port facing my cat too and here in the display filter I just want to capture STP frames which is basically an instruction to wire Wireshark to capture only the BPD use so going back to terminal now I'm going to enable the debug I'm going to run a very very simple debug once a debug spanning-tree events what I want to do here or what I'm going to try to do now keep in mind that this is timer sensitive timer sensitive in a sense that when the port initially come up both cat one and cut two are going to send their own BPD use what I want to try to do here is I want to try to catch cat to sending the proposal as well as cat one sending the proposal now if I don't see that happening I might need to shut the port down and do the know shut one or twice before I actually can catch it in that state so wish me luck here so the ports are now shut down let's see the state on cat one here I want to see administer TV done so I'm just going to bring the port up on cat one because I want to do the no shut on cat two side so now what I'm seeing is that interface is down down which is good on cat - I'm going to say no shutdown and let's see what happens this is very good now I have this in the state that I wanted to have it in the first thing that I am going to observe here is this message that says initializing portfast Ignat 24 now this simply means that the switch is preparing this port for spanning tree operations the next two lines here are telling us that cat 2 is now treating fast it in at 24 as a designated port and we can see it was actually transmitted the proposal message however we know that cat 1 is the root because we have configured it to be so so when we receive the proposal from the other side basically what we are going to do is we are going to update the rules and we are going to change this to become a root port now remember in rapid spanning tree root port transitions to forwarding state immediately the next thing that we are going to do is we are going to generate the TC trap which basically means we are going to generate the topology change notification for this port and more importantly we are going to transmit an agreement as the response to the proposal we received and at this stage the port has already changed state 2 up so we can see here in the debug this proposal and agreement going on now let's see the same thing in our Wireshark so I'm going to first stop the capture here because I have captured enough packets or frames I should say and let's focus on what is going on here now this is going to be a little bit maybe confusing because a very very similar addressing scheme that we have now we see that both switches 1 and to actually have the last octet as 9a so we have to figure out which one is which here now remember what was the first thing that we've seen the first thing that we've seen was a proposal being transmitted so we can say that this 17 5 a 9 a that this is actually cat 2 and we can say that this switch that we are seeing with 0 C 5 9a that this is our cat one so let's see how this communication went so if we take a look at the frame we're going to see here that obviously we have some bite on the wire we have the ethernet frame the logical in control and obviously we have the spanning tree bpdu where we are going to find the information about proposals and agreements are is going to be in this part that talks about bpdu Flags proposals and agreements are not separate messages they are not support kind of frames they're just simple flags in the normal configuration bpdu exchange between two switches so here we can see that one of these flags means that we are sending a proposal and if this bit is set to one this means that this is a proposal message so this was the proposal here sent by our cat - how do I know it's cat - well I can see here if I scroll this up just a little bit I'm going to see here that the bridge identifier talks about the priority of 32768 this is the default number and cat - had the default priority so I can see here that cat 2 is sending the proposal now let's see here what happened in this frame now this is the frame sent by cat 1 how can I say because I can see here that the bridge identifier is 0 which means it's cat 1 because I set the bridge priority to 0 on cat 1 and here we also see that the proposal message has been sent now let's take a look at what cat two did in response take a look the proposal is now no but what is set to one is the agreement so cat two immediately agreed that this is how the topology should look so we can see that this topology changed this topology agreement proposals and and agreements and happened relatively quickly now let me show you that in the terminal one more time so what I'm going to do here now is I'm going to go to cat one and I'm going to shut the port down mind you I have turned I have not turned on any debugs on cat one so what I'm going to do next is I'm going to turn off all the debugs on cat - and what I'm going to say is here no shutdown and I'm going to say to show spanning tree basically I can see here that the port is designate the forwarding and here on cat - I can see that the port is route forwarding how long time did this take it happened almost immediately now we know it wasn't immediately because of this proposal agreement negotiation but it happened much much faster than it would have happened in the traditional spanning tree which we have seen about five minutes ago remember how long time it took we had to transition from the listening phase and the learning phase and the falling phase on a non rapid side and in the rapid side it was the blocking discarding phase then the learning phase and only then we could move into forwarding which was controlled by the the forward delay timers in this case there are no timers involved because we are going to send our agreement after we have performed the synchronization process but keep in mind that in our case here the synchronization process was very very simple on cat - there were no other ports that were up what happens when we add the third switch I'm going to add cat 3 to the mix here by connecting it directly to cat 2 and this is going to be connected on fast hitting at 0 20 on both sides I'm not going to change any configuration that pertains to route election so our cat 1 is going to remain the route here but what I am going to change here is the configuration of my span now I'm going to modify it to actually get the traffic from this port here and send it to wire because I want to see these proposals and agreements as they are communicated between cat to and cat 3 I'm also going to keep on running my debugs here on cat too and I'm going to enable similar debugs on cat tree so let's go ahead and configure this stuff so I have cat 3 here and what I'm going to do is I'm going to change the spanning tree mode to rapid spanning tree I'm going to change spanning tree villain one priority 60 1 4 4 0 I'm going to make sure that it's not going to become the root and interface Fastnet 0 20 and I'm going to say switch port mode access and I'm going to say no shutdown here okay - I'm just going to prepare fastener 0 20 actually it was already up so let me just - shut it down here switch port mode access and I have shut it down and I'm going to say no monitor session 1 I have removed the session one I'm going to say monitor session one source interface first minute 20 and with a monitor session one destination interface fast material 3 encapsulation replicate so I'm going to create my Spanish just to make sure that fasting 0 3 is still up I'm going to make no shutdown there and I'm going to go to interface past minute 20 and I'm going to prepare for no shut down there on cat 3 I'm going to say debug panel 3 events and I'm going to do something on cat two to debug spanning tree events so now I'm ready to do no shut down here after I configure my capture here let me start it from scratch so I'm capturing traffic again from cat - and I'm going to say no shutdown let's see what happens here so now fasting at 20 is going to transition is going to become a designated port we are transmitting a proposal we have received an agreement and that's pretty much it now from cat trees perspective remember we are going to transmit the proposal we are going to update the role and we're going to transition fastener 20 to become the new route port and here in my captures I should be seeing pretty much the same thing I can see let me out stop the captures to prevent this from scrolling what I should be seeing here is that the proposal was sent we can see that the proposal was sent and then we can see that the proposal was received here we see that the proposal message has been received and here we are finally sending our own agreement all the subsequent BPD messages are not going to have proposal or agreement bits set here so this is how it works only add the third switch into the mix but I don't really want to leave it here I want to show you one more thing now at the beginning of this talk I mentioned that part of the synchronization process is that the ports are going to be moved into the disabled State then all those things are going to happen and remember what those things were so we are going to send the we are going to move the port's to discarding State and we are going to send the proposal message on downstream switches but if I take a look at the debug on my cat to know where does it actually say that any ports were moved to disabled State now the reason for that is that that happens really really quickly really really fast and I have been actually in the meantime before I recorded the previous tiny little segment of this video and now which is about I would say 25 minutes later I have been trying to actually catch it with the show commands and I realize that there is no way I'm going to be that quick and let me show you why there was no way to be that quick so what I'm going to do now is I'm going to shut the faster than a 24 down so on can't do this is the port facing our cat 1 so I'm going to bring this border I'm going to get some debug output here from from my current debug and I'm going to say under by goal because I want to run a different debug so this is going to be debug spanning-tree switch and what I want to see here are the port state changes so I'm going to run a different debug message here and let me go back here interface fastener 24 and I'm going to say no shut so let's see what happens now so there we go now take a look at this I have moved faster than a 24 from one state to another so let's ignore what was happening with fast near 24 let's focus on what was happening with fast internet 20 now you see that here at this point in time there was a blocking request for one villain one we won because we are dealing only with VLAN one now this happened here at whatever time it was at point eight zero three of a second now take a look at here this is happening less than 20 milliseconds later so it's given in 17 milliseconds later what we are seeing here is that the port changes state from blocking to forwarding because we have sent our agreement the stuff has happened on the other side and we have received sorry we have sent our we have sent out the proposal to the other side the other side our cat we perform what it needed to do and send back the agreement it happened so so fast that you cannot actually possibly catch it in this short period of time here we can see here that at the same time we were actually receiving this blocking request we transition fastly 24 240 so this is happy this has happened this moving into the discarding stake and transitioning back to 40 but it was so so quick or I should say so rapid that there was basically no impact on the forwarding traffic last but not least let's examine what happens when there is a topology change in a rapid PTSD Network first of all the topology change is a little bit redefined from the traditional spanning tree in the rapid spanning tree the topology change event is basically an on edge port transitioning to forwarding state so if the port goes down we don't really care about it we only care if the port starts forwarding and it's an on edge port when that happens two separate things happen now I'm going to talk about the last thing first just like in the traditional spanning tree we are going to send out the TC NB PDUs and they're going to propagate the same way they propagate in the traditional spanning tree this is used for backwards compatibility in a case that we have switches in our network that don't understand the rapid spanning tree process the other thing that happens is much more interesting the BPD use just like ordinary configuration BPD use are set with the TC flag set to 1 now this is sent out of every non edge port when a switch receives this bpdu what it's going to do it's going to flush it scam address table on all designated ports on all non edge designated ports and it's going to forward this BP new on similarly to the way synchronization process happens so we can think of an ever expanding wave of these VP deals with the TC bit set to 1 also this is going to happen from the originating point from the switch where the topology change actually happened for the duration of 2 hello timers so by default we are going to be observing this 2 times so we 2 hellos are going to be serviced to BP news are going to be sent this way because this is the default timer for of 2 seconds so this is very very simple you can see that it is slightly redefined but if we understand the synchronization process the proposal sent and and agreement we are going to understand how the topology change is propagated to the rapid Network

Channel: CCIEORDIE.COM

Views: 1,045

Keywords:

Id: dDNOJMmOjDU

Length: 31min 14sec (1874 seconds)

Published: Thu Feb 08 2018