Cisco CCNA ICND2 640-816

Switch STP: Enhancements to STP

by Jeremy Cioara

Start your 7-day free trial today.

This video is only available to subscribers.

A free trial includes:

  • Unlimited 24/7 access to our entire IT training video library.
  • Ability to train on the go with our mobile website and iOS/Android apps.
  • Note-taking, bookmarking, speed control, and closed captioning features.
Video Title Duration

Review: Rebuilding the Small Office Network, Part 1

Review: Rebuilding the Small Office Network, Part 2

Review: Rebuilding the Small Office Network, Part 3

Switch VLANs: Understanding VLANs

Switch VLANs: Understanding Trunks and VTP

Switch VLANs: Configuring VLANs and VTP, Part 1

Switch VLANs: Configuring VLANs and VTP, Part 2

Switch STP: Understanding the Spanning-Tree Protocol

Switch STP: Configuring Basic STP

Switch STP: Enhancements to STP

00:00:00 - As with all network technologies, as Spanning Tree has been
00:00:03 - used over the years it has evolved and it has enhanced. So what we're
00:00:08 - going to do now is take a look at the modern versions of Spanning
00:00:11 - Tree and some of the enhancements to the Spanning Tree process.
00:00:14 - First thing we'll start with is the Spanning Tree protocol port transitioning
00:00:18 - process. Meaning, why does it take so long for Spanning Tree to
00:00:22 - make a port go active. We'll then look at the initial enhancements
00:00:26 - to Spanning Tree that CISCO introduced, which is a per-VLAN
00:00:29 - Spanning Tree instance. That was the initial one that was introduced
00:00:33 - a few years back -- quite a few years back -- to allow Spanning Tree
00:00:36 - to optimize your network. We'll then look at the newest and
00:00:41 - ultimate enhancement to Spanning Tree, which is known as the
00:00:44 - rapid Spanning Tree protocol, allowing your network to converge
00:00:47 - much faster. We'll also demonstrate some of the features that
00:00:52 - rapid Spanning Tree does, and we'll try the network failure
00:00:54 - that we did in the previous video with rapid Spanning Tree
00:00:57 - enabled and see how long it takes to fail over.
00:01:01 - Well if you watched the previous video on Spanning Tree
00:01:04 - configuration, the original Spanning Tree, I think you realize
00:01:08 - that Spanning Tree has some problems. Well it's not so much
00:01:12 - problems as it is, it was just created a long time ago. When
00:01:17 - things weren't expected to move so fast, when people could
00:01:21 - it and enjoy a cup of coffee and shake your hand and look you
00:01:23 - in the eye and have a conversation. Not so nowadays. People just
00:01:27 - walk right by and how you doing, okay, great, on their way they go. And
00:01:30 - the same way with networks. Networks are expected to diverge like
00:01:34 - that, speed of light. If there's a problem, fix it; and fix
00:01:38 - it fast enough that nobody notices. So here's the problem with Spanning
00:01:42 - Tree. In its original creation, it went through two individual
00:01:46 - phases before it would actually start forwarding. You'll notice this
00:01:50 - on every single CISCO switch, when you pull it out of the box
00:01:52 - and plug a device in, the little light above the port will
00:01:56 - stay orange or that amber color until about 30 seconds
00:02:00 - go by. The initial 15 seconds, the switch is going through
00:02:05 - the listening phase. And all it's doing during those
00:02:08 - 15 seconds is listening for BPDUs. this which
00:02:12 - thought process: it thinks if you plug a device in, that
00:02:16 - device could very well be another switch that could cause a
00:02:19 - loop in the network. It might have some redundant connections
00:02:21 - I don't know about yet. So it's going to wait 15 seconds,
00:02:25 - listening to that port to see if it sees another switch's
00:02:28 - language, those BPDUs coming back through. If it does and this
00:02:33 - sport is not allowing BPDUs, it'll shut the port down. But if 15
00:02:37 - seconds goes by and no BPDUs are received, it will
00:02:41 - transition to the learning phase. Now you might
00:02:45 - remember when we were doing the previous video on configuring
00:02:48 - Spanning Tree and I caused the failure, I shut down one of the
00:02:51 - ports, you can actually watch the port. When you do the show
00:02:55 - Spanning Tree command, it's showing you the port status going
00:02:58 - through this listening or LIS is how it abbreviates it in
00:03:01 - the switch. It then transitions to learning, where it's trying
00:03:04 - to learn the MAC address that is on that port. See, if it makes
00:03:09 - the port active without knowing what MAC address is actually
00:03:12 - on that port, it's going to be very inefficient because it's
00:03:16 - going to have to start forwarding all of the packets everywhere
00:03:18 - since it doesn't really know what is on that port. So it takes
00:03:22 - another 15 seconds to learn what MAC addresses are on
00:03:27 - there. Fifteen seconds is more than enough time for a PC or
00:03:31 - a server or whatever you have plugged in to send at least one
00:03:34 - packet with its source MAC address on there and allow the switch
00:03:37 - to populate the CAM table. So 30 seconds later, the port
00:03:41 - finally transitions to forwarding and that's where the light changes
00:03:44 - from amber to green. Now that's going to cause a lot of problems
00:03:48 - in networks because -- well, I'll talk about those in just a
00:03:51 - moment. I want to talk down here about the blocking phase. I call
00:03:56 - this the bonus timer. When you have a port, let's say
00:04:00 - that you have -- let's do this, you've got a switch connected to
00:04:04 - another switch connected to another switch, just like our topology
00:04:06 - we've been working with, and we'll say that these
00:04:09 - two are active right here. This is the blocked one. Well as soon
00:04:13 - as you shut this down, the switch will implement a bonus time
00:04:18 - it's not really called that, I just call it that -- where it waits about
00:04:21 - 20 seconds before it moves a blocked port into a listening
00:04:25 - phase. What it's waiting for is it wants to see if this is coming
00:04:30 - back on line. Have you heard of a flapping interface? That's
00:04:33 - what it's waiting for. When that port goes off, meaning that
00:04:37 - it becomes disabled right here, well Spanning Tree's gonna go,
00:04:40 - well before I transition everything over, let's just wait to
00:04:43 - see if it comes back on line. I mean that could be just an administrator
00:04:47 - unplugging the port to look at the cable and plugging it back
00:04:49 - in. There is no need to upset the whole network if that port
00:04:52 - is just going to get plugged right back in. So you can see a bonus timer
00:04:56 - of up to 20 seconds, it could be a little less, but it won't
00:05:00 - go above 20, will be implemented whenever there's a failure.
00:05:03 - So what does that mean? That means that if this fails, this
00:05:08 - port before it will go active will first go through its blocking
00:05:11 - time down here, then transition to a listening state, then the
00:05:16 - transition to a learning state to learn the MAC address and then
00:05:19 - move on over to forwarding. That is up to 50 seconds of downtime
00:05:24 - per port before it can go active.
00:05:27 - These Spanning Tree delays cause two problems in our network.
00:05:31 - Number one is the problem with PCs. Modern PCs can boot faster
00:05:36 - than 30 seconds. Meaning if you get the killer laptop or
00:05:40 - brand new desktop, it's going to be able to out-boot
00:05:43 - the timer for Spanning Tree. So the port won't be active when
00:05:48 - the laptop is ready to go. Now it seems like a simple promise, like
00:05:51 - well, wait 30 seconds before you surf the internet or
00:05:54 - something, right. Well the problem is that when these PCs
00:05:57 - are booting up, they're sending out DHCP requests to get an IP address
00:06:02 - most of the time. So as they're sending out DHCP requests
00:06:05 - not getting your reply so in business now work most people
00:06:08 - are using Windows XP Professional or Windows Vista Business,
00:06:12 - and you get the little press Control Alt Delete to log on. The user
00:06:16 - goes to log on to their PC, username, type that in, password,
00:06:21 - type that in and hit Enter, and it says, sorry, the domain controllers
00:06:25 - not available, or I could not contact the domain controller
00:06:28 - Because the computer doesn't have an IP address yet. Usually
00:06:31 - when the DHCP request fails, the computer will just kind of
00:06:34 - go in this standby state where it's going to still trying and get an IP address,
00:06:37 - but it'll just wait 30 seconds or so to send out another
00:06:40 - request for an IP address. So the PCs are not able to log on to
00:06:44 - the network. The solution to that issue is portfast. If you
00:06:50 - played with CISCO's switches before, chances are you've heard
00:06:53 - of portfast. This is the cool that essentially disable spanning
00:06:58 - Tree. It turns it off on the port. And I went ahead and we'll
00:07:02 - type this in on some live switches, but I wanted to show
00:07:05 - you the warning message that you get when you type this command
00:07:08 - underneath the port. It's like a little essay that it gives you in the
00:07:12 - IOS. It says, Warning: it should not be unenabled -- or should be
00:07:16 - enabled only on ports connected to a single host. Connecting
00:07:19 - hubs, concentrators, switches, bridges, et cetera to this interface
00:07:23 - when portfast is enabled, can cause temporary bridging loops. Use with
00:07:27 - caution. So that's the warning to you, is that you are turning
00:07:31 - off Spanning Tree on this port. So the port will go active right
00:07:35 - away as soon as you plug in the device, but that could cause
00:07:38 - loops in the network. The other issue that Spanning Tree causes
00:07:43 - is that it has problems with uplink ports, meaning ports connecting
00:07:46 - to other switches. Fifty seconds of downtime in any networking
00:07:50 - cause big problems. There's a lot that can go wrong during that. The
00:07:53 - solution to that issue is the new version of Spanning Tree,
00:07:57 - Rapid Spanning Tree.
00:08:00 - Now before we get into what Rapid Spanning Tree is all about, let's
00:08:04 - first talk about the initial enhancement that CISCO made it
00:08:07 - to the normal Spanning Tree protocol. What CISCO created was
00:08:11 - a different version called PVST, that's per-VLAN Spanning
00:08:16 - Tree plus, that's CISCO's little enhancement to it to say, per-VLAN Spanning
00:08:21 - Tree by default when they first created it was only support
00:08:23 - on CISCO switches. Now it's only -- now it's become essentially
00:08:27 - an industry standard that anybody can do. But what CISCO's
00:08:30 - initial enhancement did to Spanning Tree is allow you to run
00:08:33 - an instance of Spanning Tree per VLAN.
00:08:39 - As if Spanning Tree wasn't confusing enough already, there's going
00:08:43 - to be many instances of Spanning Tree running. Here is the
00:08:46 - idea behind it. Down here I have our little topology that
00:08:49 - we're using in our network with three switches. Now in our network
00:08:53 - diagram, this one right here became the root bridge. Now ignore
00:08:57 - my little notes over here for just a moment. But just imagine
00:09:00 - that is the root bridge for everybody. What will happen is these
00:09:04 - will be the active links that are used between the two switches.
00:09:07 - This one ends up being disabled until it's needed because of
00:09:10 - a network failure. So what that means is everybody's going
00:09:13 - through the root to communicate as it should be, right. But if
00:09:17 - you think about it, you just disabled a potentially useful link in your
00:09:22 - network. Meaning this is going to be completely unused,
00:09:26 - but it would be handy if we could use it for at least some of our
00:09:29 - network traffic.
00:09:32 - That's what per-VLAN Spanning Tree is all about. What we
00:09:36 - can do is have separate Spanning Tree topologies for each
00:09:41 - VLAN that we're creating. For example, you can see my notes
00:09:45 - it allows different root bridges per VLAN. Essentially
00:09:48 - I can say this top switch is the root for VLAN 10 traffic.
00:09:53 - And what that means is for VLAN 10, these are going to
00:09:56 - be the active links and those will be used and no VLAN
00:09:59 - 10 traffic will ever cross this link over here. But I can
00:10:02 - set a separate root, thus the per VLAN Spanning Tree
00:10:07 - for other VLANs. So I can say the root for VLAN 20 is this switch
00:10:11 - right here, and if we did that the new topology looks like
00:10:15 - this. These are the -- I should -- hang on, let me do this.
00:10:20 - These are the active ports for VLAN 20 and this
00:10:25 - one ends up becoming disabled so VLAN 20 will never crosses
00:10:28 - link. What you're doing is almost a manual system of load
00:10:33 - balancing, so that VLAN 20 will start using this link
00:10:36 - that was typically unused before and still is on the phone
00:10:40 - and VLAN 10 will be using this link which is completely
00:10:43 - unused for VLAN 20. So now you've got 100 mega bits
00:10:47 - per second of dedicated bandwidth -- excuse me -- on that line for VLAN
00:10:50 - 20 and you've got 100 mega bits per second of dedicated
00:10:53 - bandwidth on that line for VLAN 10. You can do that for all
00:10:58 - the VLANs in your network. Now you can start seeing why in
00:11:01 - large networks -- man, can this be complex. You essentially
00:11:06 - can draw a separate diagram for every single VLAN that
00:11:10 - you could have. I mean, if you were drawing this out I could
00:11:13 - say, okay, well this is my VLAN 10 topology and highlight my
00:11:16 - active links for VLAN 10. I could then draw another picture
00:11:19 - over here, you know, these are my VLAN 20 topology,
00:11:23 - where we have active links right here and this one's disabled.
00:11:26 - And you could create a separate network diagram for every
00:11:29 - VLAN that you have. As a matter of fact, let me show you what this looks
00:11:32 - like on the live switches.
00:11:35 - We'll bring up my connection to switch one, and there we are.
00:11:39 - Now if you remember from the previous video when I did the initial
00:11:41 - Spanning Tree configuration, the command was show Spanning
00:11:45 - Tree and you just hit Enter. And during that video -- I have a confession
00:11:48 - to make -- I wasn't showing you the whole story. Whenever I saw this
00:11:53 - more symbol right there I just hit Q on the keyboard and say, see, look, this
00:11:56 - is how Spanning Tree's working. That's because in this video, I want
00:12:00 - to show you the rest of it. I'll hit the show Spanning Tree and
00:12:03 - you can see VLAN 1, we have the root ID, this is the
00:12:07 - root, this is switch one and so on. But watch what happens
00:12:10 - if I hit Space.
00:12:13 - Well look at this. It says for VLAN 10 I am not the root.
00:12:17 - There's another root out there that is -- my guess is
00:12:21 - that it's switch two, because that was the original root, because
00:12:25 - when we were adjusting the priority, we said that that or with
00:12:30 - the command we used was spanning tree and then I would type
00:12:33 - in VLAN 1 because I said we're using VLAN 1
00:12:36 - everywhere, which adjusted the priority for this as the
00:12:39 - root bridge on VLAN 1, but didn't adjust it for VLAN 10.
00:12:44 - and VLAN 20 and 30 and all these other
00:12:48 - ones that we're using. Look at this, VLAN 1, VLAN 10, that's
00:12:52 - I think our sales VLAN and you can see all the information
00:12:55 - there. VLAN 20, all the information there. VLAN 30,
00:12:59 - all the information right there. You can see it's
00:13:02 - the only VLAN that this is a root on is VLAN 1 and
00:13:05 - that's all VLANs we have. So by default, the CISCO switch is
00:13:09 - running per-VLAN Spanning Tree plus. So it's running an instance
00:13:13 - on every
00:13:14 - VLAN. So if you wanted to set this to be the root bridge
00:13:18 - on all the VLANs, you need to type in Spanning Tree VLAN and
00:13:21 - you can see question mark. I can type in a range of VLANs.
00:13:25 - I was just typing in one, but we can also say one, 10, 20, 30,
00:13:30 - and then do a question mark and do, you know, this will be
00:13:33 - the root primary for all of them. So now this is the true root for
00:13:38 - all the active VLANs that we have. If I do a show Spanning Tree
00:13:42 - again, now you can see VLAN 1, this bridge is the root,
00:13:46 - VLAN 10, this bridge is the root, VLAN 20, this bridge
00:13:50 - the root, and so on and so forth. So that is how you can set
00:13:54 - the switch to be the root for different VLANs and that
00:13:57 - is a nice enhancement because we can then manually set up different
00:14:01 - roots so that we use a load balancing characteristics. Some
00:14:05 - VLANs will use some links for their traffic; other VLANs
00:14:09 - will use other links. The only way you're going to be able to
00:14:12 - make this work efficiently is to have an accurate, up-to-date
00:14:17 - network diagram of your switch connections. So that you're going
00:14:21 - to be able to identify which switch ports get blocked which
00:14:24 - ones are active, and that's one of the skills that you'll have
00:14:27 - to master is based on the root bridge in the network, where
00:14:31 - the root is, which ports are actively being used. Now I say you have
00:14:35 - to master that not only for the real world kids are going to
00:14:38 - be setting this up, but be prepared when you get to certification
00:14:41 - exams to see network diagrams, to see where a root bridge is, to
00:14:45 - see the speeds of all the links, and to say, well these ports are going
00:14:48 - to be active, this one won't be active, so that will be blocked.
00:14:51 - I mean, seriously, the exams have been enhanced
00:14:56 - in a major way where you're going to be able to have a fold
00:14:59 - topology identifiying each one of those active links. It's pretty
00:15:03 - powerful. So per-VLAN Spanning Tree is an enhancement, but it's
00:15:08 - still the same old Spanning Tree engine just on multiple the
00:15:11 - So the people have spoken, the network industry has progressed
00:15:16 - and needs more speed, and you know, our signs are saying and
00:15:19 - Spanning Tree, 50 seconds, we can't have this, you know, and all
00:15:22 - the revolt. So the industry giants that create all the standards
00:15:27 - responded with a new standard, 802.1w, or the common
00:15:31 - name is Rapid Spanning Tree protocol. What it is is an enhanced
00:15:35 - version of Spanning Tree that is much more proactive than the
00:15:39 - previous. Now let me define proactive. In Spanning Tree protocol,
00:15:44 - when it finds all the active links and says, okay, these are
00:15:48 - our active links, let's block these other ones, it essentially
00:15:51 - forgets about them. Meaning they're blocked, they're not causing
00:15:55 - the loop, we're active, we're working along. So when the active link
00:15:59 - fails, Spanning Tree goes into a reactive state, meaning oh no,
00:16:04 - primarily link's lost, now what we do. You know, start looking at all
00:16:06 - these other ports to try and discover a backup path. Rapid Spanning
00:16:10 - tree, on the other hand, is proactive in the sense that once
00:16:13 - it finds its active ports, it sees the backup ports is just
00:16:18 - that -- backup.
00:16:20 - Meaning Rapid Spanning Tree remembers, if you've got this switch
00:16:24 - topology here of our three switches,
00:16:27 - and it says, these top two are going to be our primary ports, and
00:16:31 - this one will be blocked, Rapid Spanning Tree remembers, oh yeah,
00:16:36 - that can be a good backup port. I mean, it sounds so simple that
00:16:39 - that's all it really does. Spanning Tree forgets about
00:16:43 - it so if one of these dies it has to rediscover where the slowing
00:16:46 - goes. So with Rapid Spanning Tree, you get redefined port roles instead
00:16:50 - of saying it's just blocked, it will actually see it as an alternate
00:16:54 - port. The catch of Rapid Spanning Tree: it's kind of one of
00:16:58 - those things where people think oh, that's a no-brainer, let's
00:17:01 - use Rapid Spanning Tree. The catch with Rapid Spanning Tree is
00:17:05 - it's a fairly new standard. And when I say new I mean, you know,
00:17:09 - the standard -- now I'm just guesstimating here -- but the standard came
00:17:12 - out probably five years ago from now, which is 2007
00:17:18 - 2008 timeframe. So I'm talking at the end of the year, so I'm sure this
00:17:23 - recording will bleed over into the 2008. So you know, around 2003,
00:17:28 - but there are still 100 megabit per second switches is out there
00:17:31 - that work perfectly fine that
00:17:33 - you know, are five years old and people are still using them.
00:17:36 - like lies just because the standard comes out five six years
00:17:39 - ago doesn't mean everybody's like, implement that standard. So switches really
00:17:43 - only started, you know, everybody supporting Rapid Spanning Tree
00:17:46 - within the last three to four years. So in order to have Rapid
00:17:51 - Spanning Tree work, you have to have it running everywhere. Which
00:17:56 - means that there are some major network upgrades going on. If there's
00:18:00 - one switch in the network still running Spanning Tree, it will
00:18:04 - cause everybody to slow down. Because they created wrapping
00:18:07 - Spanning Tree to be backwards compatible. So if it
00:18:10 - sees an old Spanning Tree switch, it's going to say, well let's
00:18:13 - match those timers and slow ourselves down so that we can
00:18:17 - you know, work with this network. So in order to run Rapid
00:18:19 - Spanning Tree truly, it must be everywhere.
00:18:23 - I just a little bit ahead of myself and drawing a little switch
00:18:26 - network on the previous slide, but this is the official
00:18:29 - how RSTP does what it does. The main difference is just
00:18:34 - like Spanning Tree -- the old version -- it has reports it
00:18:37 - has designated ports. Meaning one per link it will have a designated
00:18:41 - port. But instead of saying it has blocked
00:18:44 - it sees those as alternate ports. That's the new port type.
00:18:48 - So instead of just saying, you're blocked, you're done, it sees this as an
00:18:52 - alternate should one of the root ports fail and the
00:18:56 - network and it can use that quickly as an alternate link. So
00:19:00 - what I want to do to wrap this video up is demonstrate the difference
00:19:04 - in speed when you compare Rapid Spanning Tree versus Spanning
00:19:08 - Tree, and I'll also show you how to turn it all on.
00:19:11 - So here's our network diagram that we've been using to test
00:19:15 - Spanning Tree. Again, what we've done is installed the redundant
00:19:18 - link connection right here on 0/24 on both
00:19:22 - of these switches. The switch one up top is our root bridge
00:19:26 - for all VLANs, and by the way CISCO does implement per-VLAN
00:19:29 - Rapid Spanning Tree. So they took their enhancement
00:19:32 - and moved into a per-VLAN level for Rapid Spanning Tree. So let's
00:19:36 - set it up, and again we still have these two hosts that can do our ping
00:19:40 - testing. We've got switch one that we're on right now,
00:19:44 - and to start Rapid Spanning Tree it's just one command. Piece
00:19:48 - of cake. Global config mode, you type in Spanning Tree mode,
00:19:53 - and then you choose which mode you want. By default, every
00:19:56 - CISCO switch runs per-VLAN Spanning Tree,
00:20:00 - but they also support multiple Spanning Tree, which is the old
00:20:03 - version, the really old version. Meaning you can run one instance
00:20:07 - of Spanning Tree per all VLANs before CISCO enhanced it. They still
00:20:12 - support that version because if you use per-VLAN Spanning
00:20:18 - Tree and have a ton of VLANs, it can actually cause
00:20:21 - a lot more processor cycles than CISCO originally intended, so
00:20:24 - you might just say, well I want to go back to multiple Spanning
00:20:27 - Tree mode where you have one Spanning Tree instance for multiple
00:20:30 - VLANs and that way you don't eat up all the resources
00:20:33 - on your switch. But what we're doing right now is converting
00:20:36 - to Rapid Spanning Tree and so you just type in Spanning Tree
00:20:39 - mode, Rapid - or Rapid PVST is the one CISCO supports. Now in
00:20:46 - order for that to work, you notice it kind of bounces the
00:20:49 - VLAN interface, in order for that to work we have to turn it on on all
00:20:52 - our switches because otherwise the timers won't increase,
00:20:56 - they'll detect Spanning Tree still running in the network. So
00:20:59 - we'll say Spanning Tree mode,
00:21:02 - Rapid on switch two. Jump over to switch three, Spanning Tree
00:21:07 - mode, Rapid. So now we're running Rapid per-VLAN Spanning
00:21:13 - Tree on all of our switches.
00:21:16 - It will take a second for them to converge. I'll just do a show Spanning
00:21:20 - Tree one more time and you can see Spanning Tree enabled, the
00:21:22 - protocol is Rapid Spanning Tree Protocol. This is still the
00:21:26 - root bridge. All the other spanning tree concepts still apply
00:21:29 - it's just now we have an -- let's go to switch three,
00:21:34 - show Spanning Tree, you see that we have this alternate port.
00:21:38 - that is labeled as that role and it's actually being used as an alternate
00:21:41 - port. It's still blocked, sure, because we still have to stop the
00:21:44 - loops, but we have now this active alternate port. And
00:21:49 - So this can still be combined with portfast to allow ports
00:21:53 - to transition quickly. So with that in place, let's
00:21:56 - do this. I'm going to open a command prompt on my computer
00:22:03 - and once again, if we go back to the network diagram, I am sitting
00:22:06 - on this computer right here, the 1.50, and I'm going to issue a
00:22:10 - continual ping over here to this PC, 1.20.
00:22:15 - And I'm going to introduce the same networks failure that
00:22:19 - we had previously. I'll ping
00:22:23 - keep that thing going. We can see that the
00:22:27 - ping is being successful. I'll scrunch this one up the job
00:22:31 - so we can watch what happens as the ping messages fail in the
00:22:34 - squish this one down right here. I'm going to jump over to
00:22:39 - switch number one, which since this is the root for all the
00:22:43 - VLANs, these are the active links that are going. This one down
00:22:46 - here is blocked. So I'm going to cause the same failure and shut down fast
00:22:50 - ethernet 0/12 and let's see how long it takes the
00:22:53 - PC to recover.
00:22:56 - So let me bring up that, bring up my continual ping going on, jump
00:23:00 - over to switch one. And i'll go into interface fast season
00:23:06 - 0/12 and you can see this thing is still
00:23:10 - going, every now and then you get an equals sign that you can watch.
00:23:13 - And I'll type in shutdown.
00:23:21 - So we have the interface went down, we have requests time out.
00:23:27 - It's supposed to be faster than that, hang on. I'm like, okay,
00:23:32 - we're failing, let me do a show Spanning Tree.
00:23:39 - We've got Rapid Spanning Tree is enabled, let me hop over to
00:23:43 - switch three. Wow, we just -- okay, wow. Yeah, that those
00:23:49 - rapid right there. That wasn't rapid at all. So let me do a show
00:23:53 - Spanning Tree on switch three. It looks like it's got the new
00:23:57 - root port that it's using. Everything's in a forwarding state.
00:24:05 - Sometimes if you cause the failure too quickly after converting
00:24:09 - to Rapid Spanning Tree, it hasn't actually detected the whole network
00:24:12 - running Rapid Spanning Tree. So let's try this again. And when they go
00:24:16 - back to switch one interface fast ethernet 0/12
00:24:19 - and I'll do a no shutdown.
00:24:22 - Power that active port back on. Now if you remember in the
00:24:25 - previous video, the no shutdown caused an immediate failure.
00:24:28 - And in the same sense here we're going to get an immediate
00:24:33 - failure. Okay, this isn't working like it's supposed to. Let's take a
00:24:37 - look. I'm going to jump over to switch three,
00:24:44 - do a show Spanning Tree. Let's take a look. It looks like we've got the
00:24:48 - designated ports that have gone into a learning state. You know what
00:24:52 - You know what I just realized?
00:24:56 - We have not turned on portfast. And when I say we, I mean I
00:24:59 - haven't. So the hosts that are pinging each other, as soon as the
00:25:02 - active link goes down, you can see that fast ethernet 0/4
00:25:06 - and eight, those are actual hosts that are plugged into that. And
00:25:10 - You notice what mode they were in -- they were in the learning mode.
00:25:13 - So I still need to turn on portfast -- I didn't even think of that that -- in order
00:25:17 - for this to take effect. So as of right now I have the original
00:25:22 - topology restored. We have a blocking port. Let me go under on
00:25:25 - switch two
00:25:29 - fast ethernet 0/8, and I'll type in Spanning Tree portfast. And
00:25:36 - there's the message, I showed you that at the beginning of this video,
00:25:38 - saying, you know, portfast is, you know, not to be enabled
00:25:43 - on non-switchboards, so I'll go under the fast ethernet 0/4, that's my router
00:25:48 - connection, also turn that on portfast, and it'll let me hop over to switch
00:25:52 - two. If I do a show IP interface
00:25:56 - I can see that I still have that PC plugged in the fast
00:25:58 - ethernet 0/8 right there. So global config, Spanning
00:26:03 - Tree portfast.
00:26:07 - Here's our little message. Let me just jump back here and mention what's happening. Whenever
00:26:11 - there is a failure in Spanning Tree and we cause this catastrophic
00:26:15 - failure, it will reset the ports. Now the ports that are portfast
00:26:18 - don't even notice that because they'll instantly go to a forwarding
00:26:22 - state. I forgot to turn on portfast on our host ports and also
00:26:26 - the router connection, that's our fast ethernet 0/4, that should be down
00:26:29 - here. So that the router, since the router is an active post
00:26:33 - it's going to go to an immediate forwarding state as well. Sell
00:26:36 - let's try that one more time.
00:26:39 - Bring up my prompt, bring up my ping message. Now I'm going to come
00:26:45 - back up to switch one, I'm going to do the same thing.
00:26:48 - I'm going to kill the fast ethernet 0/12 so the redundant
00:26:52 - link, 0/24, has to be used. Interface fast
00:26:56 - ethernet 0/12,
00:26:59 - Let me get my ping messages going back up. All right, they're pinging. And I do a
00:27:06 - shutdown.
00:27:10 - Do you see that? No, you didn't. The ping is still going.
00:27:18 - there was no failure that that's what I expected with rapids
00:27:21 - Spanning Tree. Now I'm going to jump back over to switch three and do a show
00:27:25 - Spanning Tree and you can see that the fast ethernet 0/24
00:27:28 - immediately transitioned over. That's the Spanning
00:27:32 - Tree effect I was looking for. Because as soon
00:27:35 - as it goes into the
00:27:39 - failure state it says, oh, that's my alternate poor let's
00:27:42 - immediately click over. Now when the original link gets restored,
00:27:46 - usually there's a moment to fail over let's see if it happens
00:27:49 - now. I'm going to do a no shutdown, bring that original link back online
00:27:53 - You can see the pings are still going up top.
00:28:01 - There we go, fast ethernet 0/12 has changed up. There we go, we have one ping
00:28:08 - dropped ping. That's the typical fail over time for Rapid Spanning
00:28:12 - Tree. The initial failure is immediate, meaning it's got the alternate port,
00:28:16 - click, it's shifted right back over. Then the consequent
00:28:21 - failures after that will
00:28:24 - you know, if we're switching back to the primary, will cause just
00:28:27 - a single instance, one or two seconds, of a network
00:28:30 - outage and that was enough to drop a ping packet. So
00:28:34 - overall, Rapid Spanning Tree is much faster than Spanning Tree
00:28:38 - See, we demonstrated both in one demonstration. Spanning
00:28:41 - Tree the original, you saw the 30 second timeout, and then Rapid
00:28:44 - Spanning Tree after I enabled portfast on our PC ports to
00:28:48 - take over the network and recover quickly.
00:28:52 - So Rapid Spanning Tree is designed to have almost an instantaneous
00:28:56 - failover in our network environments so that when network
00:28:59 - failures do happen, hopefully no one will even notice. So the
00:29:03 - enhancements to Spanning Tree while we walk through this, we
00:29:06 - saw the normal Spanning Tree port transition process which
00:29:09 - is the listening, learning, followed by the forwarding state, which
00:29:14 - causes a typical 30 seconds network outage or network up
00:29:18 - time while it's waiting to get to forwarding. That causes problems
00:29:22 - because PCs can boot faster than that and networks need
00:29:26 - to converge faster than that. So CISCO's initial enhancement
00:29:29 - to Spanning Tree was per-VLAN Spanning Tree, which allows
00:29:33 - you to run one instance per VLAN having different root
00:29:36 - bridges in the network, and setting up a type of manual load
00:29:40 - balancing. Finally, the ultimate enhancement to Spanning Tree is that
00:29:44 - 802.1w, or Rapid Spanning Tree, which you just saw, is
00:29:48 - a near instantaneous failover. I hope this has been informative for you and I'd like to
00:29:52 - thank you for viewing.

General Switching: Troubleshooting and Security Best Practices

Subnetting: Understanding VLSM

Routing Protocols: Distance Vector vs. Link State

Routing Protocols: OSPF Concepts

Routing Protocols: OSPF Configuration and Troubleshooting

Routing Protocols: EIGRP Concepts and Configuration

Access-Lists: The Rules of the ACL

Access-Lists: Configuring ACLs

Access-Lists: Configuring ACLs, Part 2

NAT: Understanding the Three Styles of NAT

NAT: Command-line NAT Configuration

WAN Connections: Concepts of VPN Technology

WAN Connections: Implementing PPP Authentication

WAN Connections: Understanding Frame Relay

WAN Connections: Configuring Frame Relay

IPv6: Understanding Basic Concepts and Addressing

IPv6: Configuring, Routing, and Interoperating

Certification: Some Last Words for Test Takers

Advanced TCP/IP: Working with Binary

Advanced TCP/IP: IP Subnetting, Part 1

Advanced TCP/IP: IP Subnetting, Part 2

Advanced TCP/IP: IP Subnetting, Part 3

Please help us improve by sharing your feedback on training courses and videos. For customer service questions, please contact our support team. The views expressed in comments reflect those of the author and not of CBT Nuggets. We reserve the right to remove comments that do not adhere to our community standards.

comments powered by Disqus

Course Features

Speed Control

Play videos at a faster or slower pace.


Pick up where you left off watching a video.


Jot down information to refer back to at a later time.

Closed Captions

Follow what the trainers are saying with ease.

Offline Training

Our mobile apps offer the ability to download videos and train anytime, anywhere offline.

Accountability Coaching

Develop and maintain a study plan with assistance from coaches.
Jeremy Cioara

Jeremy Cioara

CBT Nuggets Trainer

Cisco CCNA, CCDA, CCNA Security, CCNA Voice, CCNP, CCSP, CCVP, CCDP, CCIE R&S; Amazon Web Services CSA; Microsoft MCP, MCSE, Novell CNA, CNE; CompTIA A+, Network+, iNet+

Area Of Expertise:
Cisco network administration and development. Author or coauthor of numerous books, including: CCNA Voice 640-461 Official Cert Guide; CCNA Voice Official Exam Certification Guide (640-460 IIUC); CCENT Exam Prep (Exam 640-822); CCNA Exam Cram (Exam 640-802) 3rd Edition; and CCNA Voice 640-461 Official Cert Guide.

Stay Connected

Get the latest updates on the subjects you choose.

  © 2015 CBT Nuggets. All rights reserved. Licensing Agreement | Billing Agreement | Privacy Policy | RSS