WEBVTT

0:00:02.900000 --> 0:00:09.080000
 Hello and welcome to this video titled
 rapid PVST topology manipulation.

0:00:09.080000 --> 0:00:10.700000
 And what's the topic
 of this video?

0:00:10.700000 --> 0:00:13.620000
 Hey, it's manipulating
 RSTP topologies.

0:00:13.620000 --> 0:00:17.120000
 Pretty simple. So, let's
 get right into it here.

0:00:17.120000 --> 0:00:21.260000
 So, if you've done any time studying
 rapid spanning tree, you know the

0:00:21.260000 --> 0:00:25.880000
 overall goal of rapid spanning tree
 is to remove redundancy from your

0:00:25.880000 --> 0:00:30.040000
 topology. In other words, to take a
 redundant topology that has multiple

0:00:30.040000 --> 0:00:36.380000
 ways that point A could reach point B
 and strip all those ways away until

0:00:36.380000 --> 0:00:40.040000
 there's just one path between
 point A and point B.

0:00:40.040000 --> 0:00:44.140000
 And rapid spanning tree does that by
 blocking certain interfaces, placing

0:00:44.140000 --> 0:00:47.300000
 them into what's called
 the discarding state.

0:00:47.300000 --> 0:00:51.520000
 Well, that's all well and good from
 the perspective of loop prevention.

0:00:51.520000 --> 0:00:54.880000
 We want to stop those loops from circling
 around so we trim off those

0:00:54.880000 --> 0:00:55.860000
 redundant paths.

0:00:55.860000 --> 0:00:59.100000
 We're done. We have accomplished
 our mission.

0:00:59.100000 --> 0:01:01.600000
 But there is a potential
 downside to this.

0:01:01.600000 --> 0:01:06.340000
 And this downside really only involves
 the situation when a switch is

0:01:06.340000 --> 0:01:09.040000
 working with two or more VLANs.

0:01:09.040000 --> 0:01:13.020000
 Now, if a switch, if all the ports in
 that switch and all the frames going

0:01:13.020000 --> 0:01:16.860000
 through that switch are just in one
 single VLAN, you can skip this video

0:01:16.860000 --> 0:01:19.440000
 and move on. It'll be irrelevant.

0:01:19.440000 --> 0:01:24.060000
 But if you have two or more VLANs, then
 you might have a situation where

0:01:24.060000 --> 0:01:27.340000
 you need to do what I'm going to
 show you how to do right here.

0:01:27.340000 --> 0:01:31.300000
 So as you probably know, most switches
 have at least a couple of VLANs

0:01:31.300000 --> 0:01:33.980000
 in them, maybe more.

0:01:33.980000 --> 0:01:38.540000
 And typically, the connections between
 switches are configured as VLAN

0:01:38.540000 --> 0:01:42.480000
 trunks, which means they're carrying
 all the traffic from all the VLANs.

0:01:42.480000 --> 0:01:45.500000
 Now of course, you can manipulate that
 and change that, but that is the

0:01:45.500000 --> 0:01:46.920000
 default behavior.

0:01:46.920000 --> 0:01:52.780000
 Now, we know that in Cisco switches,
 Cisco switches run per VLAN rapid

0:01:52.780000 --> 0:01:56.600000
 spanning tree. In other words, each
 VLAN is creating its own separate

0:01:56.600000 --> 0:02:00.860000
 tree with its own root bridge, its
 own set of forwarding and blocking

0:02:00.860000 --> 0:02:04.920000
 ports. So if you've got 15 VLANs that
 that switch is aware of and keeping

0:02:04.920000 --> 0:02:08.500000
 track of, he's keeping track
 of 15 different trees.

0:02:08.500000 --> 0:02:10.040000
 But here's the thing.

0:02:10.040000 --> 0:02:14.440000
 If you leave spanning tree to its defaults,
 whatever switch has elected

0:02:14.440000 --> 0:02:17.180000
 the root bridge, well, let's think
 about this for a second.

0:02:17.180000 --> 0:02:20.360000
 If you don't configure spanning tree
 and you don't manipulate it, why

0:02:20.360000 --> 0:02:23.060000
 was that switch elected
 as the root bridge?

0:02:23.060000 --> 0:02:26.720000
 After all, you didn't change the bridge
 ID, you didn't modify the priority

0:02:26.720000 --> 0:02:31.260000
 in any way. So if all my switches in
 my topology, regardless of how big

0:02:31.260000 --> 0:02:37.480000
 or how small they are, all have the same
 priority of 326768, why did that

0:02:37.480000 --> 0:02:39.700000
 guy over there become
 my root bridge?

0:02:39.700000 --> 0:02:43.960000
 Well, it's because he has
 the lowest MAC address.

0:02:43.960000 --> 0:02:48.900000
 And guess what? That MAC address that
 he's using for his BPDUs and VLAN1

0:02:48.900000 --> 0:02:53.220000
 is the exact same MAC address he's
 going to use for VLAN2, VLAN3, and

0:02:53.220000 --> 0:02:57.340000
 VLAN4. So that guy over there, wherever
 he is, is going to end up being

0:02:57.340000 --> 0:03:03.580000
 the root for all of my trees, which means
 every single tree that is created

0:03:03.580000 --> 0:03:05.860000
 is going to look identical.

0:03:05.860000 --> 0:03:09.140000
 In other words, if VLAN1 has a path that's
 forwarding this way and blocking

0:03:09.140000 --> 0:03:10.500000
 this way, guess what?

0:03:10.500000 --> 0:03:14.080000
 That's exactly what VLAN2 and that's
 exactly what VLAN3 is going to look

0:03:14.080000 --> 0:03:18.160000
 like. So if I'm a switch and right
 now I have two uplinks towards the

0:03:18.160000 --> 0:03:23.880000
 root bridge, this one's forwarding,
 then that means all the traffic I

0:03:23.880000 --> 0:03:27.200000
 get, so all traffic coming in downstream
 from me, from clients or maybe

0:03:27.200000 --> 0:03:28.880000
 from other switches, guess what?

0:03:28.880000 --> 0:03:32.280000
 It's all going to go this way because
 this is the only forwarding link,

0:03:32.280000 --> 0:03:36.540000
 whether it's VLAN1, VLAN2, VLAN5, this
 is the only one that's open because

0:03:36.540000 --> 0:03:39.480000
 all the trees will look identical.


0:03:39.480000 --> 0:03:42.300000
 That can cause a problem,
 as this says right here.

0:03:42.300000 --> 0:03:45.340000
 That can cause congestion
 of links.

0:03:45.340000 --> 0:03:49.340000
 If I'm a switch and I've got 30 VLANs
 I'm keeping track of, now I've got

0:03:49.340000 --> 0:03:53.020000
 two uplinks, but really I'm only using
 this one because all the VLANs

0:03:53.020000 --> 0:03:56.580000
 from 1 to 30 are forwarding here and
 blocking here, that means I've got

0:03:56.580000 --> 0:04:00.700000
 perfectly good bandwidth on this
 link that I'm not using at all.

0:04:00.700000 --> 0:04:04.020000
 So all the traffic I'm getting is being
 pumped out this one link and if

0:04:04.020000 --> 0:04:07.940000
 this one link doesn't have enough bandwidth
 to handle all that traffic,

0:04:07.940000 --> 0:04:09.180000
 what's going to happen?

0:04:09.180000 --> 0:04:11.580000
 I'm going to have to start dropping
 frames because I'm going to have no

0:04:11.580000 --> 0:04:15.160000
 room for them. This is a situation
 where you might want to manipulate

0:04:15.160000 --> 0:04:17.420000
 spanning tree. You might want
 to say, you know what?

0:04:17.420000 --> 0:04:20.260000
 I've got some perfectly good
 bandwidth right here.

0:04:20.260000 --> 0:04:24.680000
 Why don't I take some of my VLANs and change
 some of their values artificially

0:04:24.680000 --> 0:04:29.660000
 so that maybe half the VLANs form a
 tree going this way and maybe the

0:04:29.660000 --> 0:04:33.760000
 other half of the VLANs form a tree
 going this way so I can load balance

0:04:33.760000 --> 0:04:36.780000
 my traffic and spread it
 out a little bit more.

0:04:36.780000 --> 0:04:40.680000
 It might not end up taking the most
 optimal path, but hey, if you have

0:04:40.680000 --> 0:04:43.380000
 to choose, if you've got traffic
 you say, well, I've got a choice.

0:04:43.380000 --> 0:04:47.100000
 I can either direct it along this non
-optimal path, you know, get to its

0:04:47.100000 --> 0:04:50.840000
 destination or I can send it along this
 path where it's going to be dropped

0:04:50.840000 --> 0:04:52.680000
 because there's congestion.

0:04:52.680000 --> 0:04:54.880000
 Which do you think is the
 most logical choice?

0:04:54.880000 --> 0:04:58.800000
 Let's redirect it to a non-optimal path
 so we can at least get it to where

0:04:58.800000 --> 0:05:01.520000
 it needs to go. And that's what
 we're going to talk about here.

0:05:01.520000 --> 0:05:03.260000
 How do you do that?

0:05:03.260000 --> 0:05:07.300000
 Well, at a real high level,
 there's two basic ways.

0:05:07.300000 --> 0:05:09.500000
 Now first of all, you have
 to think about this.

0:05:09.500000 --> 0:05:14.620000
 If I have a link here, okay, that's
 currently blocking, I can't use it.

0:05:14.620000 --> 0:05:17.000000
 That means one side or
 the other is blocking.

0:05:17.000000 --> 0:05:18.400000
 Well, what do we know?

0:05:18.400000 --> 0:05:21.160000
 If I have a link, let's just
 draw this right here.

0:05:21.160000 --> 0:05:27.380000
 Let's just say this is a link between
 two switches and we know that one

0:05:27.380000 --> 0:05:31.840000
 side is blocking or in rapid spanning tree
 terminology we call that discarding.

0:05:31.840000 --> 0:05:34.900000
 What do we know about
 the other side?

0:05:34.900000 --> 0:05:38.660000
 Don't we know that the other
 side has to be forwarding?

0:05:38.660000 --> 0:05:41.860000
 It's got to be a designated port.

0:05:41.860000 --> 0:05:45.200000
 That's the only combination
 you could have.

0:05:45.200000 --> 0:05:49.160000
 A blocking port will never lead to
 a root port on the other side and a

0:05:49.160000 --> 0:05:53.040000
 blocking port will never lead to another
 blocking port on the other side.

0:05:53.040000 --> 0:05:56.320000
 The only way this guy on the left is
 blocking is because every two seconds

0:05:56.320000 --> 0:06:02.020000
 he's receiving BPDUs and something about
 those BPDUs has told him, hey,

0:06:02.020000 --> 0:06:03.280000
 you need to block.

0:06:03.280000 --> 0:06:05.860000
 He's learned that this port right here,
 let's just say it's fast, Ethan,

0:06:05.860000 --> 0:06:09.860000
 at zero slash two, he's learned due to
 these BPDUs that that is not worthy

0:06:09.860000 --> 0:06:12.100000
 of being his root port.

0:06:12.100000 --> 0:06:16.860000
 He has another port somewhere else
 that's his root port that's better.

0:06:16.860000 --> 0:06:21.760000
 And he's also learned that this device
 up here is actually closer to the

0:06:21.760000 --> 0:06:24.560000
 root bridge than he is.

0:06:24.560000 --> 0:06:27.140000
 So he's got no choice
 but to go blocking.

0:06:27.140000 --> 0:06:29.280000
 All right. So that
 being considered.

0:06:29.280000 --> 0:06:31.640000
 Now I have associated with the situation
 where I say, well, hold on a

0:06:31.640000 --> 0:06:34.720000
 second. Yeah, that's all well and good,
 but I need to change that behavior

0:06:34.720000 --> 0:06:39.000000
 because I need to send
 traffic up that link.

0:06:39.000000 --> 0:06:42.960000
 Well, if I'm going to send traffic up
 that link, then that means I have

0:06:42.960000 --> 0:06:44.900000
 to have both sides
 forwarding, right?

0:06:44.900000 --> 0:06:47.460000
 That's the only way you can get
 traffic across that link.

0:06:47.460000 --> 0:06:53.600000
 Well, if you have a link and both sides
 are going to be forwarding, once

0:06:53.600000 --> 0:06:55.540000
 again, what do you know?

0:06:55.540000 --> 0:06:58.260000
 You know that one side has
 to be a designated port.

0:06:58.260000 --> 0:07:02.500000
 One guy has to be responsible for
 delivering the BPDUs on the wire.

0:07:02.500000 --> 0:07:05.100000
 The other side has to
 receive the BPDUs.

0:07:05.100000 --> 0:07:07.320000
 He has to be a root port.

0:07:07.320000 --> 0:07:09.980000
 You're never going to have a situation
 where both sides are forwarding

0:07:09.980000 --> 0:07:13.780000
 and both sides are designated
 or both sides are root.

0:07:13.780000 --> 0:07:15.060000
 That would be illegal.

0:07:15.060000 --> 0:07:16.480000
 That would break spanning trees.

0:07:16.480000 --> 0:07:19.420000
 So if you got a forwarding link, one
 side is designated and one side is

0:07:19.420000 --> 0:07:21.480000
 root. So let's think about this.

0:07:21.480000 --> 0:07:23.980000
 Here I have a situation
 with two links.

0:07:23.980000 --> 0:07:29.320000
 Okay. And we know that right now everything
 left to their defaults.

0:07:29.320000 --> 0:07:30.780000
 One side is forwarding.

0:07:30.780000 --> 0:07:31.920000
 One side is blocking.

0:07:31.920000 --> 0:07:34.840000
 So we know this forwarding side
 has to be a designated port.

0:07:34.840000 --> 0:07:36.460000
 It's the only thing he can be.

0:07:36.460000 --> 0:07:40.300000
 And then we have another link that's
 forwarding on both sides.

0:07:40.300000 --> 0:07:43.960000
 Okay. And if I'm on this bottom switch
 right here, I can guarantee you

0:07:43.960000 --> 0:07:48.460000
 this forwarding link down below is going
 to be a root port and the forwarding

0:07:48.460000 --> 0:07:51.700000
 link up above is going to
 be a designated port.

0:07:51.700000 --> 0:07:54.920000
 Now for whatever reason, you know,
 maybe this link here on the left is

0:07:54.920000 --> 0:07:56.460000
 becoming congested.

0:07:56.460000 --> 0:08:01.180000
 Let's just write that here in
 my chicken scratch, congested.

0:08:01.180000 --> 0:08:05.180000
 So we need to start sending traffic up
 the link to the right, which means

0:08:05.180000 --> 0:08:10.320000
 we got to take this blocking port and
 change him to becoming our new root

0:08:10.320000 --> 0:08:13.240000
 port. That is our objective here.

0:08:13.240000 --> 0:08:15.900000
 If I have a link that's currently blocking
 and I want to make it forwarding

0:08:15.900000 --> 0:08:20.780000
 on both sides, that means just leave
 the side that's already forwarding

0:08:20.780000 --> 0:08:23.820000
 alone. You don't have to worry about
 him, but we got to do something on

0:08:23.820000 --> 0:08:29.360000
 this bottom link here to make him believe
 that's actually better to get

0:08:29.360000 --> 0:08:34.800000
 to the root bridge this way than what
 he currently believes, which is

0:08:34.800000 --> 0:08:38.820000
 this way. He currently has a legitimate
 reason to think on the left that

0:08:38.820000 --> 0:08:41.360000
 that is a better path
 to the root bridge.

0:08:41.360000 --> 0:08:43.340000
 We have to change that behavior
 through commands.

0:08:43.340000 --> 0:08:48.180000
 We have to say, no, I'm going to change
 something on you or maybe on the

0:08:48.180000 --> 0:08:52.640000
 switch on top of you to make you think
 it's better to go to the right.

0:08:52.640000 --> 0:08:55.500000
 That way I can make that link forwarding
 and forwarding and I can start

0:08:55.500000 --> 0:08:57.360000
 redirecting traffic for that.

0:08:57.360000 --> 0:09:02.380000
 So that's what PVST or rapid PVST
 manipulation is all about.

0:09:02.380000 --> 0:09:06.980000
 And remember, we're doing
 this on a per VLAN basis.

0:09:06.980000 --> 0:09:08.520000
 There wouldn't be much sense.

0:09:08.520000 --> 0:09:09.860000
 I mean, you certainly
 could do this.

0:09:09.860000 --> 0:09:13.020000
 You could say, hey, I just want to make
 the left link blocking and make

0:09:13.020000 --> 0:09:15.460000
 the right link forwarding
 for everything.

0:09:15.460000 --> 0:09:17.740000
 You could do that, but
 what's the point?

0:09:17.740000 --> 0:09:21.220000
 Now, instead of traffic being congested
 on the left link, now traffic

0:09:21.220000 --> 0:09:23.360000
 is going to be congested
 on the right link.

0:09:23.360000 --> 0:09:27.600000
 So we want to make these changes only
 for a subset of our traffic, only

0:09:27.600000 --> 0:09:28.760000
 for some of the VLANs.

0:09:28.760000 --> 0:09:30.880000
 So that's going to be part of your decision
 is you're doing this in the

0:09:30.880000 --> 0:09:34.660000
 lab or you're doing it in a real topology
 is you have to decide for yourself,

0:09:34.660000 --> 0:09:39.720000
 okay, some of my VLANs are going to
 take a non-optimal path just so at

0:09:39.720000 --> 0:09:42.200000
 least they're not getting dropped and
 you'll have to make the decision

0:09:42.200000 --> 0:09:45.120000
 which VLANs do I want
 to do this for.

0:09:45.120000 --> 0:09:48.120000
 So let's take a look and
 see how we would do this.

0:09:48.120000 --> 0:09:51.580000
 Let's start with this
 topology right here.

0:09:51.580000 --> 0:09:54.820000
 So notice in the upper left corner,
 so first of all, we've got two VLANs

0:09:54.820000 --> 0:09:59.500000
 we're dealing with on the bottom,
 we see VLAN 2 and VLAN 3.

0:09:59.500000 --> 0:10:02.220000
 And notice that in the upper left corner,
 we've got a router in the cloud

0:10:02.220000 --> 0:10:05.140000
 and he's serving as
 the default gateway.

0:10:05.140000 --> 0:10:11.180000
 So his link right here, whatever that
 interface is that the IP address

0:10:11.180000 --> 0:10:14.300000
 on that interface is serving as the
 default gateway for all these people

0:10:14.300000 --> 0:10:15.980000
 down here in VLAN 2.

0:10:15.980000 --> 0:10:19.120000
 And similarly, this interface right here
 is serving as the default gateway

0:10:19.120000 --> 0:10:21.200000
 for all the people in VLAN 3.

0:10:21.200000 --> 0:10:24.260000
 Now let me ask you this,
 pause for just a moment.

0:10:24.260000 --> 0:10:26.880000
 I want you to pause this video
 and figure this out.

0:10:26.880000 --> 0:10:30.980000
 What will be the forwarding and blocking
 state of all of your interfaces

0:10:30.980000 --> 0:10:35.460000
 for VLAN 2? So go ahead and pause this
 video and maybe take a screenshot

0:10:35.460000 --> 0:10:39.380000
 of this or maybe just write down a piece
 of paper and I want you to figure

0:10:39.380000 --> 0:10:43.120000
 out which ports will be forwarding
 as root ports, which ports will be

0:10:43.120000 --> 0:10:47.560000
 forwarding as designated ports and
 what will be blocking for VLAN 2 in

0:10:47.560000 --> 0:10:50.380000
 this topology. Go ahead and press play
 when you think you've figured that

0:10:50.380000 --> 0:10:58.540000
 out. Okay, so let's
 work through this.

0:10:58.540000 --> 0:10:59.780000
 Hopefully this is what
 you came up with.

0:10:59.780000 --> 0:11:02.020000
 I'm just going to do this
 really, really fast.

0:11:02.020000 --> 0:11:04.540000
 And if you're not familiar with how
 I came up with this, I'd encourage

0:11:04.540000 --> 0:11:07.860000
 you to review some other videos about
 how to come up with spanning tree

0:11:07.860000 --> 0:11:12.440000
 topologies. First we have to
 identify the root bridge.

0:11:12.440000 --> 0:11:15.060000
 Now the switch on the bottom,
 we can't really factor him in.

0:11:15.060000 --> 0:11:16.580000
 We don't have his bridge ID.

0:11:16.580000 --> 0:11:20.120000
 So we're just going to assume that
 he is not the root bridge here.

0:11:20.120000 --> 0:11:25.700000
 So they all start out
 with 3,278, 0211.

0:11:25.700000 --> 0:11:27.160000
 And then we have some change here.


0:11:27.160000 --> 0:11:30.580000
 So the switch on the top is
 going to be my root bridge.

0:11:30.580000 --> 0:11:36.240000
 So let's make him right there because
 1, 2 is lower than AB on any of

0:11:36.240000 --> 0:11:37.220000
 the other switches.

0:11:37.220000 --> 0:11:40.640000
 So that means that all of his ports
 are going to be designated ports.

0:11:40.640000 --> 0:11:44.680000
 So let's just mark that down here,
 forwarding designated port.

0:11:44.680000 --> 0:11:48.100000
 And forwarding designated port.

0:11:48.100000 --> 0:11:51.580000
 Now right now, remember, I'm just
 figuring this out for VLAN 2.

0:11:51.580000 --> 0:11:55.760000
 All right, and hopefully you figured
 out that these were my root ports

0:11:55.760000 --> 0:11:59.320000
 that are forwarding on these
 respective switches.

0:11:59.320000 --> 0:12:03.780000
 So forwarding root port.

0:12:03.780000 --> 0:12:07.320000
 And this guy over here,
 forwarding root port.

0:12:07.320000 --> 0:12:12.820000
 All right, and then the switch on the
 bottom, switch X, he has 3 equal

0:12:12.820000 --> 0:12:15.000000
 cost paths to get to
 the root bridge.

0:12:15.000000 --> 0:12:19.260000
 But the port he's going to choose for
 his root port is going to be 0 slash

0:12:19.260000 --> 0:12:23.260000
 3. And that's because of
 his upstream neighbor.

0:12:23.260000 --> 0:12:29.860000
 His upstream neighbor has
 0 slash 1 and 0 slash 2.

0:12:29.860000 --> 0:12:33.620000
 So the switch X on the bottom is going
 to prefer his neighbor's port of

0:12:33.620000 --> 0:12:39.580000
 0 slash 1. All right, and then this
 is going to be blocking right here.

0:12:39.580000 --> 0:12:45.020000
 This is going to be
 blocking right here.

0:12:45.020000 --> 0:12:49.560000
 OK, these upstream ports here are going
 to be designated ports because

0:12:49.560000 --> 0:12:52.200000
 they are closer to
 the root bridge.

0:12:52.200000 --> 0:12:54.920000
 Then switch X is.

0:12:54.920000 --> 0:12:59.900000
 And this right here is also
 going to be blocking.

0:12:59.900000 --> 0:13:06.380000
 And then of course, all these ports
 here leading to the VLAN 2 hosts,

0:13:06.380000 --> 0:13:09.080000
 every single one of those
 is going to be forwarding.

0:13:09.080000 --> 0:13:11.080000
 All right, so there
 is our topology.

0:13:11.080000 --> 0:13:15.520000
 So we can see that right now when VLAN
 2 wants to get to their default

0:13:15.520000 --> 0:13:18.680000
 gateway, there's only one path
 they could possibly take.

0:13:18.680000 --> 0:13:24.320000
 Going across here,
 up here, up here.

0:13:24.320000 --> 0:13:28.740000
 And then around and up.

0:13:28.740000 --> 0:13:33.780000
 That is how VLAN 2 is going to
 get to their default gateway.

0:13:33.780000 --> 0:13:37.920000
 OK, now if we figured out what VLAN
 3 looks like, it would be exactly

0:13:37.920000 --> 0:13:41.960000
 the same thing. The only change in
 VLAN 3 would be when we get to the

0:13:41.960000 --> 0:13:44.060000
 root bridge. That's where
 the traffic would diverge.

0:13:44.060000 --> 0:13:48.960000
 But VLAN 3 would still go up this way,
 up and around, and then it would

0:13:48.960000 --> 0:13:51.480000
 change it to go across here.

0:13:51.480000 --> 0:13:56.500000
 So we can see here we've got two potential
 places where traffic could

0:13:56.500000 --> 0:14:00.020000
 be congested. Traffic could be congested
 on this link right here because

0:14:00.020000 --> 0:14:01.860000
 it's all taking the left link.

0:14:01.860000 --> 0:14:04.700000
 Or traffic could be congested right
 here because it's all going across

0:14:04.700000 --> 0:14:09.280000
 that way. So we need to do some load
 balancing here to split this traffic

0:14:09.280000 --> 0:14:11.820000
 apart to prevent that.

0:14:11.820000 --> 0:14:16.380000
 Alright, so first of all we've got to
 pick one of the VLANs or the other.

0:14:16.380000 --> 0:14:20.600000
 So let's pick VLAN 2 since
 we worked on that.

0:14:20.600000 --> 0:14:21.940000
 Look at this topology.

0:14:21.940000 --> 0:14:26.620000
 It would make more sense if VLAN
 2 traffic followed this path.

0:14:26.620000 --> 0:14:30.980000
 If it came in here, then
 went around and up.

0:14:30.980000 --> 0:14:33.400000
 Right? That would make more sense.


0:14:33.400000 --> 0:14:36.060000
 So what we need to do is we
 need to change this link.

0:14:36.060000 --> 0:14:39.980000
 That path right there, we need to take
 this blocking port and make him

0:14:39.980000 --> 0:14:43.780000
 believe he should be a root port.

0:14:43.780000 --> 0:14:49.300000
 Well, in this topology, the easiest
 way to do that would be to split our

0:14:49.300000 --> 0:14:51.200000
 root bridges apart.

0:14:51.200000 --> 0:14:55.700000
 So what if we did something to VLAN
 2 so that instead of the switch on

0:14:55.700000 --> 0:14:59.480000
 the very top being the root bridge, what
 if we made the switch over here?

0:14:59.480000 --> 0:15:01.240000
 The root bridge.

0:15:01.240000 --> 0:15:07.160000
 That would actually solve all of our
 problems because if switch 44, 44

0:15:07.160000 --> 0:15:11.900000
 was the root, then switch X would naturally
 change this blocking port

0:15:11.900000 --> 0:15:14.560000
 into a forwarding root port.

0:15:14.560000 --> 0:15:15.780000
 And we'd be done.

0:15:15.780000 --> 0:15:18.980000
 All the traffic would
 go this direction.

0:15:18.980000 --> 0:15:23.420000
 So this is when talking about per VLAN
 load balancing with spanning tree.

0:15:23.420000 --> 0:15:25.940000
 This is the first objective
 you should have.

0:15:25.940000 --> 0:15:29.540000
 You should take a look at your topologies
 and you should ask yourself,

0:15:29.540000 --> 0:15:33.860000
 okay, after I've identified the VLAN,
 I want to start manipulating.

0:15:33.860000 --> 0:15:37.160000
 Is there a way I could take the root
 bridge for that VLAN and move it

0:15:37.160000 --> 0:15:41.560000
 somewhere else? If I can move the root
 bridge to another part of my topology,

0:15:41.560000 --> 0:15:43.600000
 will that meet my objectives?

0:15:43.600000 --> 0:15:45.560000
 Now, sometimes the
 answer will be no.

0:15:45.560000 --> 0:15:48.260000
 We'll see in the next topology
 that that won't help us.

0:15:48.260000 --> 0:15:51.420000
 But in this topology, it
 absolutely would help us.

0:15:51.420000 --> 0:15:55.220000
 So by simply making the switch on the
 left, the root bridge, and we could

0:15:55.220000 --> 0:15:58.680000
 do that by lowering his priority, that'd
 probably be the easiest thing.

0:15:58.680000 --> 0:16:03.300000
 So on the switch on the left, we would
 just type spanning dash tree, VLAN

0:16:03.300000 --> 0:16:06.460000
 2 priority, zero.

0:16:06.460000 --> 0:16:08.320000
 We would just do that on
 the switch on the left.

0:16:08.320000 --> 0:16:14.220000
 At the global level, spanning dash
 tree, VLAN 2 priority, zero.

0:16:14.220000 --> 0:16:19.440000
 And that would make switch 44, 44,
 our root bridge problem solved.

0:16:19.440000 --> 0:16:22.800000
 Now our traffic is diverging
 right down there at switch X.

0:16:22.800000 --> 0:16:25.120000
 Don't have any congested
 links anymore.

0:16:25.120000 --> 0:16:26.400000
 So that's one way.

0:16:26.400000 --> 0:16:31.000000
 But sometimes that
 method won't work.

0:16:31.000000 --> 0:16:34.080000
 For example, let's take a look
 at the next topology here.

0:16:34.080000 --> 0:16:36.520000
 Notice the one little
 change I've made.

0:16:36.520000 --> 0:16:39.740000
 We don't see the router in
 the cloud, but he's there.

0:16:39.740000 --> 0:16:42.280000
 The router in the cloud still exists
 and he's still serving as the default

0:16:42.280000 --> 0:16:47.100000
 gateway. But now there's only one way
 to get to him going across the 10

0:16:47.100000 --> 0:16:48.880000
 gigabit ethernet link.

0:16:48.880000 --> 0:16:52.680000
 So once again, this nothing
 here has changed.

0:16:52.680000 --> 0:16:55.740000
 So I'm just going to quickly sort of
 draw out with my chicken scratch.

0:16:55.740000 --> 0:16:58.960000
 You'll have to bear with me here what
 the topology is going to look like.

0:16:58.960000 --> 0:17:01.400000
 So still, this guy is still going to
 end up being the root bridge for

0:17:01.400000 --> 0:17:04.680000
 both two and three by default.

0:17:04.680000 --> 0:17:07.880000
 So he's going to be forwarding
 as designated ports.

0:17:07.880000 --> 0:17:11.300000
 These guys are going to be
 forwarding as root ports.

0:17:11.300000 --> 0:17:15.060000
 These ports will be forwarding as designated
 ports because they're closer

0:17:15.060000 --> 0:17:16.940000
 to the root bridge.

0:17:16.940000 --> 0:17:22.580000
 OK, so nothing's changed there and still
 the point on the left here will

0:17:22.580000 --> 0:17:24.440000
 be our forwarding root port.

0:17:24.440000 --> 0:17:26.220000
 This will be blocking.

0:17:26.220000 --> 0:17:28.060000
 This will be blocking.

0:17:28.060000 --> 0:17:29.260000
 This will be blocking.

0:17:29.260000 --> 0:17:32.680000
 All right. So we're right back
 to where we were again.

0:17:32.680000 --> 0:17:39.460000
 Now once again, everything left to its
 defaults, all the traffic is going

0:17:39.460000 --> 0:17:46.260000
 to come in, switch X, go up this
 path, go across and then go over.

0:17:46.260000 --> 0:17:50.320000
 Now this 10 gig link, there's
 nothing we can do about that.

0:17:50.320000 --> 0:17:52.380000
 That's the only link we have
 to get to the root bridge.

0:17:52.380000 --> 0:17:54.740000
 So there's no place to split
 across the traffic.

0:17:54.740000 --> 0:18:01.280000
 But still this link here could be congested
 and this link could be congested.

0:18:01.280000 --> 0:18:06.980000
 Now in this particular case, we could
 do the same thing as before actually

0:18:06.980000 --> 0:18:07.820000
 looking at this.

0:18:07.820000 --> 0:18:13.820000
 So if we made the switch on the left
 root bridge for VLAN 2, that would

0:18:13.820000 --> 0:18:15.000000
 solve our problem.

0:18:15.000000 --> 0:18:16.240000
 OK, I sort of lied to you.

0:18:16.240000 --> 0:18:17.700000
 I said, oh, look, this
 topology is different.

0:18:17.700000 --> 0:18:18.540000
 It's not going to work here.

0:18:18.540000 --> 0:18:20.100000
 Well, actually it will work here.

0:18:20.100000 --> 0:18:23.000000
 But I want to show you
 an alternative method.

0:18:23.000000 --> 0:18:28.080000
 So what else could we do?

0:18:28.080000 --> 0:18:31.900000
 So if our objective is the same as before,
 all right, we still have the

0:18:31.900000 --> 0:18:39.340000
 same objective, which is we want to take
 some of the traffic and redirect

0:18:39.340000 --> 0:18:41.260000
 it. Oh, let's just be fun.

0:18:41.260000 --> 0:18:44.640000
 Let's say and redirect it this
 way just to be on the fun side.

0:18:44.640000 --> 0:18:46.040000
 All right. So we want
 to send it over there.

0:18:46.040000 --> 0:18:47.940000
 You say, OK, let's make
 that the root bridge.

0:18:47.940000 --> 0:18:49.260000
 Yes, we could do that.

0:18:49.260000 --> 0:18:52.540000
 We could make 44 E for the root,
 wash our hands and walk away.

0:18:52.540000 --> 0:18:55.780000
 And honestly, that's what you
 should always try first.

0:18:55.780000 --> 0:18:57.700000
 OK, if that works, do it.

0:18:57.700000 --> 0:18:59.660000
 You're done. Walk away.

0:18:59.660000 --> 0:19:03.700000
 But let's say for whatever reason,
 it's not going to work or we want a

0:19:03.700000 --> 0:19:04.480000
 different approach.

0:19:04.480000 --> 0:19:06.820000
 OK, so our objective
 is still the same.

0:19:06.820000 --> 0:19:09.720000
 We still want to take
 0 slash 4 down here.

0:19:09.720000 --> 0:19:12.940000
 And for one of our VLANs, let's
 just focus on VLAN 2.

0:19:12.940000 --> 0:19:14.900000
 We want to make that
 the root port.

0:19:14.900000 --> 0:19:17.480000
 OK, so how do we do that?

0:19:17.480000 --> 0:19:19.400000
 Well, let's think about
 this here for a second.

0:19:19.400000 --> 0:19:22.380000
 If we're going to keep the switch on
 the top, if we're going to leave

0:19:22.380000 --> 0:19:25.820000
 him alone as the current root
 bridge for VLAN 2 and VLAN 3.

0:19:25.820000 --> 0:19:31.160000
 And yet, I want switch X to go off
 on 0 slash 4 as his root port.

0:19:31.160000 --> 0:19:32.760000
 What does that really mean?

0:19:32.760000 --> 0:19:35.400000
 That means that somehow I'm just going
 to delete all this stuff here because

0:19:35.400000 --> 0:19:36.880000
 it's going to look like
 chicken scratch.

0:19:36.880000 --> 0:19:41.340000
 Somehow I have to make switch X believe
 it's a better path this way to

0:19:41.340000 --> 0:19:46.060000
 get to the root bridge than his
 current path, which is this way.

0:19:46.060000 --> 0:19:48.320000
 How do I do that?

0:19:48.320000 --> 0:19:51.240000
 Well, let me ask you a question.

0:19:51.240000 --> 0:19:55.560000
 How does the switch determine of all
 of its ports, which port is going

0:19:55.560000 --> 0:19:58.020000
 to be the root port
 in the first place?

0:19:58.020000 --> 0:20:02.460000
 Isn't the number one
 consideration cost?

0:20:02.460000 --> 0:20:05.800000
 If a switch says, oh, I'm getting BPUs
 from the root bridge on this interface

0:20:05.800000 --> 0:20:10.080000
 and this interface isn't the very first
 thing he does is to say, what's

0:20:10.080000 --> 0:20:12.660000
 the total cost to the
 root if I go that way?

0:20:12.660000 --> 0:20:15.580000
 What's the total cost to the
 root if I go that way?

0:20:15.580000 --> 0:20:18.600000
 Well, if he's choosing this as the root
 port right here and he's blocking

0:20:18.600000 --> 0:20:23.560000
 this, it's probably not always, but it's
 probably because this has a total

0:20:23.560000 --> 0:20:27.460000
 path cost that's lower than this.

0:20:27.460000 --> 0:20:30.440000
 So isn't that something
 we could change?

0:20:30.440000 --> 0:20:34.920000
 Absolutely. So I could say, all right,
 well, if I want him to go off to

0:20:34.920000 --> 0:20:38.800000
 the right and I don't want to manipulate
 the root bridges, that would

0:20:38.800000 --> 0:20:39.720000
 be my only solution.

0:20:39.720000 --> 0:20:43.840000
 I have to somehow make him think that
 the cost, the total cost, not just

0:20:43.840000 --> 0:20:48.680000
 the cost necessarily of this link,
 but the total cost to the right is

0:20:48.680000 --> 0:20:52.580000
 lower than the total
 cost going upstream.

0:20:52.580000 --> 0:20:55.580000
 Now, in this topology, it's actually
 fairly simple to do that.

0:20:55.580000 --> 0:21:00.460000
 So we know that when the BPUs are flowing
 down this way, the initial cost

0:21:00.460000 --> 0:21:02.080000
 is going to be zero.

0:21:02.080000 --> 0:21:04.060000
 And when they're going down this way,
 the initial cost is going to be

0:21:04.060000 --> 0:21:08.640000
 zero. And then when 43.44 sends them
 down this way, he's going to say,

0:21:08.640000 --> 0:21:11.000000
 hey, my cost to the
 root bridge is four.

0:21:11.000000 --> 0:21:15.660000
 When 44.E4 is sending them down this way,
 he's going to say the same thing.

0:21:15.660000 --> 0:21:18.700000
 My cost is four.

0:21:18.700000 --> 0:21:23.640000
 And then when switch X gets it, he's
 going to add on his own local cost,

0:21:23.640000 --> 0:21:27.100000
 which is 19. Here it's 19.

0:21:27.100000 --> 0:21:30.020000
 He's going to add that to the cost that
 his neighbor is advertising, which

0:21:30.020000 --> 0:21:35.280000
 is four. And this case is going to say,
 well, my total cost going up to

0:21:35.280000 --> 0:21:40.980000
 the straight up is 23.

0:21:40.980000 --> 0:21:44.940000
 My total cost going over
 to the right is 23.

0:21:44.940000 --> 0:21:50.500000
 It's a tie. But now when he looks at
 his neighbor's bridge IDs, he's going

0:21:50.500000 --> 0:21:53.000000
 to say, well, both neighbors
 started out with 3278.

0:21:53.000000 --> 0:21:55.180000
 But here's where we have
 some difference.

0:21:55.180000 --> 0:21:59.340000
 He's going to say this neighbor is
 a little bit different because he's

0:21:59.340000 --> 0:22:05.300000
 ABCD 43. This guy is ABCD 44.

0:22:05.300000 --> 0:22:07.360000
 43 is lower than 44.

0:22:07.360000 --> 0:22:07.860000
 That's why I'm saying that he's
 going to add on his own.

0:22:07.860000 --> 0:22:10.100000
 He made this the root port.

0:22:10.100000 --> 0:22:11.980000
 So here's what we're going to do.

0:22:11.980000 --> 0:22:17.080000
 We're just going to make him believe
 that this total cost is less than

0:22:17.080000 --> 0:22:21.960000
 23. If we can make him believe it's
 22 or 21 or or anything less than

0:22:21.960000 --> 0:22:25.640000
 23, he'll prefer the
 path to the right.

0:22:25.640000 --> 0:22:28.120000
 So there's there's two places
 we could do this.

0:22:28.120000 --> 0:22:31.120000
 This is going to involve
 lowering cost.

0:22:31.120000 --> 0:22:35.340000
 We could go on to port
 0 slash 2 up here.

0:22:35.340000 --> 0:22:38.900000
 Default cost is 4 because
 that's the gig link.

0:22:38.900000 --> 0:22:43.440000
 What if we went on to that interface
 and just for VLAN 2 only, we said,

0:22:43.440000 --> 0:22:44.540000
 hey, guess what?

0:22:44.540000 --> 0:22:48.920000
 For VLAN 2, your local
 path cost is 3.

0:22:48.920000 --> 0:22:54.620000
 Now he would be advertising 3
 downstream to his neighbor.

0:22:54.620000 --> 0:22:57.180000
 That would influence switch X.

0:22:57.180000 --> 0:22:59.280000
 Switch X would say, oh,
 now my cost is 22.

0:22:59.280000 --> 0:23:02.440000
 If I go up to the right,
 which is lower than 23.

0:23:02.440000 --> 0:23:09.200000
 If I go straight up so we could do that
 or we could go to his own interface.

0:23:09.200000 --> 0:23:10.780000
 Zero slash 4 right here.

0:23:10.780000 --> 0:23:12.260000
 And we could say, hey, guess what?


0:23:12.260000 --> 0:23:13.820000
 Your cost is not 19.

0:23:13.820000 --> 0:23:16.540000
 Why don't we knock you down to 18?


0:23:16.540000 --> 0:23:19.020000
 It would have the
 exact same effect.

0:23:19.020000 --> 0:23:22.780000
 Now he would say, oh, my neighbor is
 advertising 4 to me, but I'm 18,

0:23:22.780000 --> 0:23:24.540000
 4 plus 18 is 22.

0:23:24.540000 --> 0:23:28.340000
 Either way, he would think going up to
 the right is lower cost than going

0:23:28.340000 --> 0:23:34.360000
 straight up. So in order to do that,
 we would just let's just do it right

0:23:34.360000 --> 0:23:36.560000
 here on his own local interface.

0:23:36.560000 --> 0:23:38.560000
 It's currently 19.

0:23:38.560000 --> 0:23:44.600000
 I want to knock that down to
 let's say 18 for VLAN 2 only.

0:23:44.600000 --> 0:23:50.600000
 So the command to do that is give myself
 enough room here because it's

0:23:50.600000 --> 0:23:52.380000
 a pretty long command.

0:23:52.380000 --> 0:23:55.220000
 You say, so we would
 go to the interface.

0:23:55.220000 --> 0:24:08.640000
 And on that interface, we would say,
 spanning dash tree VLAN 2 cost 18.

0:24:08.640000 --> 0:24:12.440000
 So that's how you lower the cost
 on a particular interface.

0:24:12.440000 --> 0:24:15.900000
 And it will influence the local switch
 on which you've done the command

0:24:15.900000 --> 0:24:18.620000
 and any switches beneath him.

0:24:18.620000 --> 0:24:22.720000
 So in this particular case, we probably
 would want to do it right here

0:24:22.720000 --> 0:24:27.700000
 on switch X. Now we could have
 done it on switch 4E4.

0:24:27.700000 --> 0:24:30.920000
 I could have done that exact same command
 right here instead of making

0:24:30.920000 --> 0:24:33.860000
 the cost 18. I could
 have made it 3.

0:24:33.860000 --> 0:24:35.740000
 But here's the downside to this.

0:24:35.740000 --> 0:24:40.820000
 If there's any other switches beneath
 this guy, it'll affect all of them.

0:24:40.820000 --> 0:24:46.460000
 Because now whenever switch 44E4 sends
 a BPDU downstream, it's going to

0:24:46.460000 --> 0:24:50.620000
 affect all those BPDUs because I've
 affected the cost on his upstream

0:24:50.620000 --> 0:24:55.740000
 report. Whereas if I do the change right
 here on switch X, only it only

0:24:55.740000 --> 0:24:57.060000
 affects switch X.

0:24:57.060000 --> 0:25:00.580000
 He's the last switch in the line changing
 the cost on his local interface

0:25:00.580000 --> 0:25:03.800000
 isn't going to affect
 anybody beneath him.

0:25:03.800000 --> 0:25:08.420000
 So that's one way we can manipulate
 load balancing is by changing the

0:25:08.420000 --> 0:25:12.140000
 costs of links. All right,
 here's another option.

0:25:12.140000 --> 0:25:14.820000
 Right now we know he's
 going up this way.

0:25:14.820000 --> 0:25:17.400000
 Right, this is forwarding
 designated port.

0:25:17.400000 --> 0:25:19.720000
 This is forwarding root port.

0:25:19.720000 --> 0:25:22.180000
 Well, if I said, hey, here's
 what I want to do.

0:25:22.180000 --> 0:25:25.180000
 I want to change him so
 he goes up this way.

0:25:25.180000 --> 0:25:27.660000
 Right now this is designated port.


0:25:27.660000 --> 0:25:29.140000
 This is blocking.

0:25:29.140000 --> 0:25:32.420000
 So once again, I'm just
 doing this for VLAN 2.

0:25:32.420000 --> 0:25:37.100000
 Same objective. I want to make the blocking
 port switch over to becoming

0:25:37.100000 --> 0:25:41.180000
 the new root port, which will cause this
 old root port to go into blocking.

0:25:41.180000 --> 0:25:43.040000
 Now I could change cost.

0:25:43.040000 --> 0:25:44.580000
 I could certainly do that.

0:25:44.580000 --> 0:25:49.480000
 Now in this case, if I want to change
 the cost, it wouldn't make any sense

0:25:49.480000 --> 0:25:52.400000
 to do it on zero slash
 three up above.

0:25:52.400000 --> 0:25:55.980000
 Because if I make that a cost of three,
 for example, instead of four,

0:25:55.980000 --> 0:25:59.320000
 which it is right now, that's going
 to affect both of the BPDUs that he

0:25:59.320000 --> 0:26:00.920000
 sends downstream.

0:26:00.920000 --> 0:26:05.000000
 Switch X is still going to see both
 links as being an equal cost path.

0:26:05.000000 --> 0:26:08.620000
 And he'll still default to choosing
 this link over here on the left.

0:26:08.620000 --> 0:26:10.380000
 So that doesn't make any sense.

0:26:10.380000 --> 0:26:15.760000
 So in this case case, if I want to modify
 cost, I would not do it on the

0:26:15.760000 --> 0:26:17.300000
 upstream switch.

0:26:17.300000 --> 0:26:19.320000
 I would do it on port
 zero slash two.

0:26:19.320000 --> 0:26:20.520000
 Actually, I could do
 one of two things.

0:26:20.520000 --> 0:26:23.140000
 I could either make zero
 slash two look better.

0:26:23.140000 --> 0:26:26.760000
 I could say, hey, instead
 of 19, let's make you 18.

0:26:26.760000 --> 0:26:31.160000
 Or I could go to zero slash three
 and make him look worse.

0:26:31.160000 --> 0:26:35.020000
 I could say, hey, instead
 of 19, we'll make you 20.

0:26:35.020000 --> 0:26:38.360000
 That would work.

0:26:38.360000 --> 0:26:42.300000
 Or alternatively, let's think
 about this here for a second.

0:26:42.300000 --> 0:26:48.500000
 Why did the switch on the bottom choose
 the path on the left in the first

0:26:48.500000 --> 0:26:53.060000
 place? Why did the path on the left
 become his root port instead of the

0:26:53.060000 --> 0:26:54.480000
 path on the right?

0:26:54.480000 --> 0:26:56.640000
 It wasn't due to cost, was it?

0:26:56.640000 --> 0:27:02.420000
 Originally, when both BPDUs came flowing
 down, here's one and here's one,

0:27:02.420000 --> 0:27:07.080000
 they both had a cost of four because
 that's what his upstream neighbor's

0:27:07.080000 --> 0:27:11.300000
 cost is. Right here in this link, his
 cost is four to get to the root

0:27:11.300000 --> 0:27:13.700000
 bridge. Cost was the same.

0:27:13.700000 --> 0:27:14.620000
 So what did he do?

0:27:14.620000 --> 0:27:18.980000
 Switch X went back to the BPDU and
 looked at the sending bridge ID.

0:27:18.980000 --> 0:27:22.340000
 But both BPDUs had exactly
 the same sending bridge ID.

0:27:22.340000 --> 0:27:24.640000
 They were both being
 sent from 43, 44.

0:27:24.640000 --> 0:27:26.180000
 So then what did Switch X do?

0:27:26.180000 --> 0:27:29.080000
 He went back to the BPDUs and
 he looked at the port IDs.

0:27:29.080000 --> 0:27:30.780000
 And didn't it look like this?

0:27:30.780000 --> 0:27:39.880000
 Didn't both of the port IDs have a priority,
 which was the same 128, 128.

0:27:39.880000 --> 0:27:43.420000
 But here's where our differentiating,
 our differentiating, our differing

0:27:43.420000 --> 0:27:49.960000
 factors were. The port on the left
 had a port ID of 128 and 0 slash 1.

0:27:49.960000 --> 0:27:54.680000
 The one on the right had
 128 and 0 slash 2.

0:27:54.680000 --> 0:28:01.300000
 That's why Switch X originally shows
 port 0 slash 3, because this whole

0:28:01.300000 --> 0:28:05.020000
 thing is lower than
 this whole thing.

0:28:05.020000 --> 0:28:11.380000
 Well, we can't change the 0, 1 and 0,
 2, but we can change the priority.

0:28:11.380000 --> 0:28:13.440000
 We can change the port priority.

0:28:13.440000 --> 0:28:16.800000
 So here we can go to his upstream
 neighbor, 0 slash 2.

0:28:16.800000 --> 0:28:20.760000
 We can get onto that interface so we
 can say, hey, your port priority,

0:28:20.760000 --> 0:28:27.340000
 it's not 128. Let's lower you
 down to something like 112.

0:28:27.340000 --> 0:28:31.780000
 And so now if the BPDUs that were popping
 out of that port said, hey,

0:28:31.780000 --> 0:28:37.980000
 my port priority is 112, that would
 influence his downstream neighbor,

0:28:37.980000 --> 0:28:41.980000
 Switch X. Switch X would say, oh, okay,
 well, that port is better than

0:28:41.980000 --> 0:28:43.360000
 the port on the left.

0:28:43.360000 --> 0:28:47.840000
 I'll choose my 0 slash 2 as my
 root port and go to the right.

0:28:47.840000 --> 0:28:49.460000
 So that's what we're going to do.

0:28:49.460000 --> 0:28:55.280000
 We're going to modify the port priority
 on his upstream neighbor.

0:28:55.280000 --> 0:28:57.400000
 And here's the command to do that.


0:28:57.400000 --> 0:29:04.380000
 So it's once again, an
 interface command.

0:29:04.380000 --> 0:29:08.480000
 We got to go to the interface first
 and say, spanning dash tree, VLAN

0:29:08.480000 --> 0:29:16.980000
 2 port dash priority, 1 12.

0:29:16.980000 --> 0:29:20.200000
 And we do that on the
 upstream switch.

0:29:20.200000 --> 0:29:22.440000
 And now that influences
 the downstream switch.

0:29:22.440000 --> 0:29:26.200000
 So now the port, now the BPDUs is popping
 out of here, have the default

0:29:26.200000 --> 0:29:28.360000
 port priority of 128.

0:29:28.360000 --> 0:29:32.560000
 And the BPDUs that are popping out
 here have a modified port priority

0:29:32.560000 --> 0:29:40.580000
 of 112. That will affect switch X and
 he'll choose 0 slash 2 as his root

0:29:40.580000 --> 0:29:44.540000
 port. Now one last thing I want to
 say about this before I bring this

0:29:44.540000 --> 0:29:45.860000
 video to a close.

0:29:45.860000 --> 0:29:49.560000
 You might be wondering, Keith, where
 did you get that number 112?

0:29:49.560000 --> 0:29:53.600000
 Okay, I recognize I have to lower 128
 to something lower, but why 112?

0:29:53.600000 --> 0:29:57.780000
 Why not 14 or 6 or 100?

0:29:57.780000 --> 0:30:02.880000
 Well, the reason for this is because
 you may recall in some of my other

0:30:02.880000 --> 0:30:06.020000
 videos or videos you've watched from
 other instructors when we talked

0:30:06.020000 --> 0:30:08.120000
 about bridge priorities.

0:30:08.120000 --> 0:30:10.620000
 Okay, remember when we talked
 about bridge priorities?

0:30:10.620000 --> 0:30:15.140000
 We said that originally the bridge
 priority field was a 16 bit field.

0:30:15.140000 --> 0:30:23.020000
 So here was my entire BPDU, my entire
 BPDU and somewhere in here we had

0:30:23.020000 --> 0:30:27.800000
 a 16 bit field, 12345678,
 12345678.

0:30:27.800000 --> 0:30:34.740000
 That was originally
 our bridge priority.

0:30:34.740000 --> 0:30:37.420000
 And then we added the Mac to it.

0:30:37.420000 --> 0:30:42.680000
 This whole thing became
 the bridge ID.

0:30:42.680000 --> 0:30:47.180000
 But remember bridges do things
 called extended system IDs.

0:30:47.180000 --> 0:30:51.880000
 Where really only the last
 four bits are the priority.

0:30:51.880000 --> 0:30:57.320000
 The rest of this stuff right here is
 the extended system ID, which in

0:30:57.320000 --> 0:30:58.980000
 most cases of the VLAN.

0:30:58.980000 --> 0:31:04.440000
 And so because of that, this bit right
 here, this bit right there was

0:31:04.440000 --> 0:31:07.540000
 our bit 4096. This is a review.

0:31:07.540000 --> 0:31:11.860000
 So when you were changing your bridge
 priority, it had to be an increments

0:31:11.860000 --> 0:31:14.980000
 or multiples of 4096.

0:31:14.980000 --> 0:31:17.920000
 The lowest one you could
 have would be 000, 000.

0:31:17.920000 --> 0:31:19.360000
 That was setting the bridge
 priority to zero.

0:31:19.360000 --> 0:31:25.400000
 The next one you could have would
 be 0001, which was 4096.

0:31:25.400000 --> 0:31:26.160000
 Well, guess what?

0:31:26.160000 --> 0:31:30.020000
 Port priorities are exactly
 the same way.

0:31:30.020000 --> 0:31:37.220000
 So if we take a look
 at our BPDU again.

0:31:37.220000 --> 0:31:43.780000
 And the port priority field
 is, let's see here.

0:31:43.780000 --> 0:31:47.060000
 That's going to be in
 increments of this.

0:31:47.060000 --> 0:31:54.200000
 So we've got 124816.

0:31:54.200000 --> 0:31:55.760000
 That's going to be an 8 bit field.


0:31:55.760000 --> 0:32:08.060000
 12345678. Alright, so right here, this
 is going to be my port priority.

0:32:08.060000 --> 0:32:10.260000
 So with port priority.

0:32:10.260000 --> 0:32:14.540000
 See if I can figure this out here.


0:32:14.540000 --> 0:32:18.040000
 So this is going to be 124816.

0:32:18.040000 --> 0:32:23.120000
 Only these bits here are actually
 modifiable for port priority.

0:32:23.120000 --> 0:32:27.920000
 I don't remember off the top
 of my head what went in here.

0:32:27.920000 --> 0:32:33.380000
 I don't recall that, but the important
 thing is, is that the first port

0:32:33.380000 --> 0:32:37.100000
 is the value of 16.

0:32:37.100000 --> 0:32:42.500000
 Right in binary, we've got
 124816, so on and so forth.

0:32:42.500000 --> 0:32:48.240000
 So with port priorities, your
 default value is 128.

0:32:48.240000 --> 0:32:50.780000
 So let me get rid of
 some of this here.

0:32:50.780000 --> 0:32:56.460000
 So here's your default value
 with the 128 bit turned on.

0:32:56.460000 --> 0:32:58.680000
 Everything else turned off.

0:32:58.680000 --> 0:33:03.460000
 The lowest port priority
 we can have is that.

0:33:03.460000 --> 0:33:07.160000
 But remember that right there
 is the 16 bit in binary.

0:33:07.160000 --> 0:33:11.400000
 So that means our port priorities always
 have to be in increments or multiples

0:33:11.400000 --> 0:33:17.260000
 of 16. So if I started
 with 128, which is.

0:33:17.260000 --> 0:33:23.940000
 Well, whatever that is, the
 next increment down is 112.

0:33:23.940000 --> 0:33:26.320000
 If we subtract 16 from that.

0:33:26.320000 --> 0:33:29.740000
 That gives us 112.

0:33:29.740000 --> 0:33:32.680000
 And you actually see that if you're on
 a Cisco switch, you don't necessarily

0:33:32.680000 --> 0:33:35.100000
 have to memorize that.

0:33:35.100000 --> 0:33:37.780000
 For example, I'll show you.

0:33:37.780000 --> 0:33:46.900000
 If we go to a Cisco switch.

0:33:46.900000 --> 0:34:00.640000
 I'll pick some random port here
 interface fast, Ethan at zero.

0:34:00.640000 --> 0:34:02.560000
 Zero slash eight.

0:34:02.560000 --> 0:34:06.480000
 Spanning dash tree V land
 one port dash priority.

0:34:06.480000 --> 0:34:09.880000
 If we do a question mark,
 see how it says here.

0:34:09.880000 --> 0:34:12.660000
 Port priority in increments of 16.


0:34:12.660000 --> 0:34:14.340000
 You can have zero through 240.

0:34:14.340000 --> 0:34:18.520000
 So if I put something that was not
 an increment of 16 like 100.

0:34:18.520000 --> 0:34:21.480000
 They would say, no, can't do that.


0:34:21.480000 --> 0:34:25.700000
 It has to be an increment of 16.

0:34:25.700000 --> 0:34:29.060000
 So finishing off here, going back to
 this topology we were just looking

0:34:29.060000 --> 0:34:35.360000
 at. When you have a
 situation like this.

0:34:35.360000 --> 0:34:40.180000
 Where one port is forwarding.

0:34:40.180000 --> 0:34:42.120000
 One port is blocking.

0:34:42.120000 --> 0:34:44.180000
 And you want to switch that.

0:34:44.180000 --> 0:34:47.220000
 You want to make the forwarding
 port blocking.

0:34:47.220000 --> 0:34:49.940000
 You want to make the blocking
 port forwarding.

0:34:49.940000 --> 0:34:51.340000
 And here's the key thing.

0:34:51.340000 --> 0:34:57.040000
 If both ports lead upstream
 to the same upstream switch.

0:34:57.040000 --> 0:35:01.580000
 Then you can modify port priority
 to accomplish your objectives.

0:35:01.580000 --> 0:35:04.900000
 You know the default
 port priority is 128.

0:35:04.900000 --> 0:35:09.700000
 So by simply lowering that
 in increments of 16.

0:35:09.700000 --> 0:35:13.400000
 So the next lowest
 one down is 112.

0:35:13.400000 --> 0:35:19.620000
 That will accomplish your mission and it
 will affect the downstream neighbor.

0:35:19.620000 --> 0:35:26.820000
 So that concludes this video on rapid
 per rapid per VLAN spanning tree

0:35:26.820000 --> 0:35:28.920000
 load balancing. Thank
 you for watching.

