1
00:00:00,000 --> 00:00:04,000
In a previous video, we optimize Spanning Tree

2
00:00:04,000 --> 00:00:08,000
to make switch 1 the  root for VLANs 1 and 10

3
00:00:08,000 --> 00:00:10,000
and switch 2 the root for VLAN 20.

4
00:00:10,000 --> 00:00:15,000
On switch 1 as an example sh spanning-tree vlan 1

5
00:00:15,000 --> 00:00:19,000
shows me that the switch is forwarding on all ports

6
00:00:19,000 --> 00:00:26,000
and the same is true for VLAN 10 but for VLAN 20

7
00:00:26,000 --> 00:00:30,000
the switch is blocking on gigabit 0/1.

8
00:00:30,000 --> 00:00:35,000
The same is true on switch 2 sh spanning-tree vlan 1

9
00:00:35,000 --> 00:00:40,000
switch is blocking on gigabit 0/1 for VLAN 10

10
00:00:40,000 --> 00:00:44,000
switch is blocking on gigabit 0/1

11
00:00:44,000 --> 00:00:48,000
but is forwarding on all ports for VLAN 20.

12
00:00:48,000 --> 00:00:56,000
So in other words, traffic that sent by say this PC to NPM is gonna be sent

13
00:00:56,000 --> 00:01:03,000
to switch 2 and then its gonna be forwarded across gigabit 0/0 to get to NPM.

14
00:01:03,000 --> 00:01:06,000
That’s ok if you only have a few PCs

15
00:01:06,000 --> 00:01:11,000
but if you had multiple access switches with their host in VLAN 20

16
00:01:11,000 --> 00:01:14,000
their traffic would be sent to this switch

17
00:01:14,000 --> 00:01:17,000
and traffic would have to go across this gigabit link

18
00:01:17,000 --> 00:01:23,000
which will become a bottleneck to get to say multiple servers on the left-hand side.

19
00:01:23,000 --> 00:01:28,000
So traffic that's traversing from 1 core switch to another

20
00:01:28,000 --> 00:01:31,000
is being limited to using this gigabit 0/0 link.

21
00:01:31,000 --> 00:01:37,000
So what we're going to do is we're going to bond or bind these 2 physical interfaces

22
00:01:37,000 --> 00:01:40,000
into a logical EtherChannel or link aggregation

23
00:01:40,000 --> 00:01:46,000
so that Spanning Tree sees the 2 physical ports as a single port

24
00:01:46,000 --> 00:01:48,000
and doesn’t block either of the ports.

25
00:01:48,000 --> 00:01:54,000
So once again on switch 1 notice that gigabit 0/1 is blocking

26
00:01:54,000 --> 00:02:01,000
that will change in a moment when we create our link aggregation or EtherChannel.

27
00:02:01,000 --> 00:02:04,000
So to configure an EtherChannel I’m gonna type

28
00:02:04,000 --> 00:02:11,000
conf t type interface range gigabitEthernet 0/0 - 1

29
00:02:11,000 --> 00:02:15,000
so I’m gonna make configuration changes on both of these interfaces at the same time.

30
00:02:15,000 --> 00:02:18,000
I’m gonna shut the ports down and then I’m gonna type

31
00:02:18,000 --> 00:02:28,000
switchport trunk encapsulation dot1q switchport mode trunk

32
00:02:28,000 --> 00:02:33,000
channel channel-group 1

33
00:02:33,000 --> 00:02:40,000
in other words, I’m going to put both of these interfaces into EtherChannel 1

34
00:02:40,000 --> 00:02:42,000
you can create multiple EtherChannels on a switch.

35
00:02:42,000 --> 00:02:46,000
So as an example, I could have 2 interfaces down to this access switch

36
00:02:46,000 --> 00:02:51,000
and bonded them together in link aggregation 2 or EtherChannel 2

37
00:02:51,000 --> 00:02:54,000
but in this example, I’m gonna use EtherChannel 1

38
00:02:54,000 --> 00:02:56,000
I’m going to specify a mode

39
00:02:56,000 --> 00:03:03,000
and in this example, I’m going to use LACP active mode.

40
00:03:03,000 --> 00:03:07,000
Now in EtherChannel you have 3 ways of doing this

41
00:03:07,000 --> 00:03:11,000
if you set it to on it means that you manually create an EtherChannel

42
00:03:11,000 --> 00:03:14,000
and there’s no negotiation with the other side

43
00:03:14,000 --> 00:03:22,000
the interface are simply added to an EtherChannel, you could also use LACP or PAgP

44
00:03:22,000 --> 00:03:27,000
now LACP or Link Aggregation Control Protocol is an industry standard protocol

45
00:03:27,000 --> 00:03:29,000
that allows switches to negotiate

46
00:03:29,000 --> 00:03:35,000
the forming of link aggregated ports or EtherChannels as Cisco calls it.

47
00:03:35,000 --> 00:03:41,000
Port aggregation protocol or PAgP is a Cisco propriety protocol

48
00:03:41,000 --> 00:03:45,000
That allows you to configure link aggregated ports.

49
00:03:45,000 --> 00:03:50,000
So you either set it on with no negotiation with the neighboring device.

50
00:03:50,000 --> 00:03:54,000
So switch 1 as an example, will not negotiate with switch 2

51
00:03:54,000 --> 00:04:01,000
to form a link aggregation or an EtherChannel or you specify LACP or PAgP

52
00:04:01,000 --> 00:04:04,000
now when using LACP you have 2 options.

53
00:04:04,000 --> 00:04:11,000
Active means that it will negotiate with the other side to form a link aggregation.

54
00:04:11,000 --> 00:04:16,000
So it's actively trying to form a link aggregation with the remote end.

55
00:04:16,000 --> 00:04:22,000
Passive means that the device is waiting for LACP messages neighboring device

56
00:04:22,000 --> 00:04:24,000
before it forms a link aggregation.

57
00:04:24,000 --> 00:04:27,000
So you want to ensure that you don’t set both sides to passive

58
00:04:27,000 --> 00:04:30,000
because that means that both sides are waiting

59
00:04:30,000 --> 00:04:33,000
for the other side to form the link aggregation

60
00:04:33,000 --> 00:04:38,000
and because neither of them are initiating the link aggregation it won’t be formed.

61
00:04:38,000 --> 00:04:41,000
So you can either set both sides to active or 1 side to active

62
00:04:41,000 --> 00:04:43,000
and the other side to passive

63
00:04:43,000 --> 00:04:46,000
but don’t set both sides to passive.

64
00:04:46,000 --> 00:04:49,000
In our example, we're going to set both sides to active

65
00:04:49,000 --> 00:04:52,000
now with PAgP you have a similar kind of idea.

66
00:04:52,000 --> 00:04:57,000
PAgP desirable mode means that the switches going to ask

67
00:04:57,000 --> 00:04:59,000
the other side to set up the link aggregation.

68
00:04:59,000 --> 00:05:04,000
Auto means that it will wait for the other side to initiate the link aggregation.

69
00:05:04,000 --> 00:05:06,000
So in our example

70
00:05:06,000 --> 00:05:11,000
we’ve set link aggregation or the port channel mode to active.

71
00:05:11,000 --> 00:05:16,000
So were using LACP or LACP or link aggregation

72
00:05:16,000 --> 00:05:18,000
to form a link aggregation with the remote side.

73
00:05:18,000 --> 00:05:21,000
Now from a Spanning Tree point of view

74
00:05:21,000 --> 00:05:24,000
we're going to set the link type to point-to-point

75
00:05:24,000 --> 00:05:29,000
to allow Spanning Tree to  negotiate things quicker

76
00:05:29,000 --> 00:05:32,000
we don’t want to use a shared link

77
00:05:32,000 --> 00:05:37,000
we want to use point-to-point links to enhanced the Spanning Tree convergence timers.

78
00:05:37,000 --> 00:05:41,000
In Rapid Spanning Tree if a link you shared

79
00:05:41,000 --> 00:05:44,000
in other words, the duplex is half it uses timers

80
00:05:44,000 --> 00:05:47,000
such as blocking, listening, learning and forwarding

81
00:05:47,000 --> 00:05:49,000
but if it's a point-to-point link

82
00:05:49,000 --> 00:05:52,000
Spanning Tree doesn’t have to wait for timers to expire

83
00:05:52,000 --> 00:05:54,000
for ports to start forwarding

84
00:05:54,000 --> 00:05:56,000
so you need to use point-to-point links

85
00:05:56,000 --> 00:06:00,000
if you want to make use of rapid convergence and Rapid Spanning Tree.

86
00:06:00,000 --> 00:06:03,000
So I’m going to no shut the interfaces at this port.

87
00:06:03,000 --> 00:06:07,000
Let’s do something similar on switch 2

88
00:06:07,000 --> 00:06:10,000
now you might want to wait till both sides

89
00:06:10,000 --> 00:06:13,000
are I configured before you no shut the interfaces

90
00:06:13,000 --> 00:06:16,000
because otherwise, you get messages such as the following.

91
00:06:16,000 --> 00:06:20,000
LACP is not enabled on the remote end

92
00:06:20,000 --> 00:06:24,000
so my link aggregation or EtherChannel hasn’t been formed,

93
00:06:24,000 --> 00:06:32,000
so let’s look at the output of the show EtherChannel summary command

94
00:06:32,000 --> 00:06:40,000
Notice at the moment we have 2 ports that have been added to EtherChannel 1

95
00:06:40,000 --> 00:06:46,000
we're using the LACP protocol but notice D means that the ports down.

96
00:06:46,000 --> 00:06:52,000
So we are setting up a layer 2 link aggregation, not a layer 3

97
00:06:52,000 --> 00:06:55,000
so we are using switching rather than routing.

98
00:06:55,000 --> 00:06:59,000
So we'll use trunks across this ports as an example

99
00:06:59,000 --> 00:07:06,000
but the ports are currently down because the switch cannot negotiate with the other end.

100
00:07:06,000 --> 00:07:08,000
The link aggregation is not working.

101
00:07:08,000 --> 00:07:11,000
sh run will show me my configuration

102
00:07:11,000 --> 00:07:15,000
notice there's my port channel and on my 2 interfaces

103
00:07:15,000 --> 00:07:20,000
notice these 2 interfaces are part of the link aggregation or port channel

104
00:07:20,000 --> 00:07:22,000
which is port channel 1

105
00:07:22,000 --> 00:07:30,000
we're using LACP it’s a trunk port so using a layer 2 interface

106
00:07:30,000 --> 00:07:35,000
and we're running Spanning Tree across this making them point to point links.

107
00:07:35,000 --> 00:07:40,000
That’s means that Spanning Tree will converge quicker.

108
00:07:40,000 --> 00:07:44,000
sh etherchannel port-channel

109
00:07:44,000 --> 00:07:48,000
shows me some additional information

110
00:07:48,000 --> 00:07:52,000
such as the number of ports in the link aggregation being 0

111
00:07:52,000 --> 00:07:56,000
no ports are currently in this port channel.

112
00:07:56,000 --> 00:08:00,000
So let’s configure switch 2 and see if it makes any difference.

113
00:08:00,000 --> 00:08:09,000
So here switch 2, conf t int range gigabitEthernet 0/0 - 1

114
00:08:09,000 --> 00:08:14,000
so these 2 core ports shut the interfaces down.
e

115
00:08:14,000 --> 00:08:28,000
switchport trunk encapsulation dot1q switchport mode trunk channel-group

116
00:08:28,000 --> 00:08:32,000
we're going to use the same number in this case so EtherChannel 1

117
00:08:32,000 --> 00:08:34,000
doesn’t have to be the same on both sides.

118
00:08:34,000 --> 00:08:40,000
The mode we're going to use is active because we want to use LACP

119
00:08:40,000 --> 00:08:46,000
and we want this ports to also initiate link aggregation with remote side.

120
00:08:46,000 --> 00:08:51,000
spanning-tree link-type point-to-point

121
00:08:51,000 --> 00:08:53,000
sh run

122
00:08:53,000 --> 00:08:57,000
let’s see what we’ve configured and then I'll no shut the interface.

123
00:08:57,000 --> 00:09:00,000
So there's our port channel or EtherChannel.

124
00:09:00,000 --> 00:09:05,000
Here’s our configuration on the first port

125
00:09:05,000 --> 00:09:08,000
and configuration on the second port.

126
00:09:08,000 --> 00:09:12,000
Now something that's really important is make sure

127
00:09:12,000 --> 00:09:16,000
that the configuration on all ports is the same.

128
00:09:16,000 --> 00:09:19,000
So in other words, the speed and the duplex

129
00:09:19,000 --> 00:09:23,000
the switch type all your settings need to be the same on both sides

130
00:09:23,000 --> 00:09:27,000
to guarantee that the link aggregation comes up.

131
00:09:27,000 --> 00:09:34,000
So I’ve no shut the port or rather no shut the port

132
00:09:34,000 --> 00:09:38,000
let’s see if the negotiation takes place correctly.

133
00:09:38,000 --> 00:09:41,000
So we can see the interface came up

134
00:09:41,000 --> 00:09:45,000
there’s gigabit 0/0 coming up, here’s gigabit 0/1 coming up

135
00:09:45,000 --> 00:09:48,000
so line status has changed to up

136
00:09:48,000 --> 00:09:57,000
sh etherchannel summary, our 2 parts are bundled in a port channel.

137
00:09:57,000 --> 00:10:03,000
So P using the protocol LACP, it’s a layer 2 EtherChannel.

138
00:10:03,000 --> 00:10:08,000
So S for layer 2 and the ports are in used. So that's good.

139
00:10:08,000 --> 00:10:11,000
Let’s look at the port channel

140
00:10:11,000 --> 00:10:19,000
so we see more information now than we saw on switch 1 previously.

141
00:10:19,000 --> 00:10:22,000
So our port channel is port channel 1

142
00:10:22,000 --> 00:10:25,000
the number of ports in the channel are 2

143
00:10:25,000 --> 00:10:30,000
we can see the 2 ports that are active in the channel

144
00:10:30,000 --> 00:10:33,000
are gigabit 0/0 and gigabit 0/1

145
00:10:33,000 --> 00:10:38,000
the last port that was bundled is 0/1, so back on switch 1

146
00:10:38,000 --> 00:10:43,000
previously notice no ports were part of the port channel

147
00:10:43,000 --> 00:10:46,000
but here the port channel did come up.

148
00:10:46,000 --> 00:10:49,000
So if we do the command again

149
00:10:49,000 --> 00:10:53,000
we now see that 2 ports are in the port channel

150
00:10:53,000 --> 00:10:55,000
or link aggregation or EtherChannel

151
00:10:55,000 --> 00:11:00,000
and both ports gigabit 0/0 and 0/1 are active.

152
00:11:00,000 --> 00:11:03,000
So we’ve now bundled these 2 ports in a link aggregation

153
00:11:03,000 --> 00:11:05,000
how does that affect Spanning Tree?

154
00:11:05,000 --> 00:11:08,000
so sh spanning-tree vlan 20

155
00:11:08,000 --> 00:11:13,000
Notice VLAN 20 is forwarding on all ports on switch 1

156
00:11:13,000 --> 00:11:20,000
Vlan 1 all ports are forwarding, vlan 10 all ports are forwarding.

157
00:11:20,000 --> 00:11:25,000
Now previously on switch 1 gigabit 0/1 was blocking

158
00:11:25,000 --> 00:11:28,000
but notice now all ports are forwarding.

159
00:11:28,000 --> 00:11:34,000
The root port of switch 1 is the port channel or EtherChannel.

160
00:11:34,000 --> 00:11:38,000
Notice the path cost has gone down from 4 to 3

161
00:11:38,000 --> 00:11:40,000
because it sees this as a better path.

162
00:11:40,000 --> 00:11:43,000
Port is forwarding it’s the root port.

163
00:11:43,000 --> 00:11:50,000
On switch 2 sh spanning-tree vlan 1

164
00:11:50,000 --> 00:11:52,000
all ports are forwarding

165
00:11:52,000 --> 00:11:56,000
port channel 1 or EtherChannel 1 is the root port

166
00:11:56,000 --> 00:12:00,000
and were forwarding, cost has also gone down to 3

167
00:12:00,000 --> 00:12:04,000
same for VLAN 10 and for VLAN 20.

168
00:12:04,000 --> 00:12:07,000
All ports are forwarding on this core switches.

169
00:12:07,000 --> 00:12:11,000
So for redundancy and for higher throughput

170
00:12:11,000 --> 00:12:15,000
between our core switches, we want to enable link aggregation

171
00:12:15,000 --> 00:12:18,000
or EtherChannel also called port channels.

172
00:12:18,000 --> 00:12:22,000
Link aggregation or the bonding of multiple  ports

173
00:12:22,000 --> 00:12:26,000
will give us the ability to load balance traffic across the link aggregation.

174
00:12:26,000 --> 00:12:31,000
It also means that Spanning Tree is not blocking 1 of the ports

175
00:12:31,000 --> 00:12:34,000
so we get much better throughput across the links.

176
00:12:34,000 --> 00:12:38,000
It also provides redundancy because if one of the ports goes down

177
00:12:38,000 --> 00:12:40,000
the port channel will still be up.

178
00:12:40,000 --> 00:12:46,000
So as an example, for VLAN 20 on switch 1 the root port is port 1

179
00:12:46,000 --> 00:12:50,000
if I go into gigabit 0/0 and shut the port down

180
00:12:50,000 --> 00:12:56,000
and then type show etherchannel summary

181
00:12:56,000 --> 00:13:00,000
we can see that one of the ports has gone down

182
00:13:00,000 --> 00:13:03,000
but one of the port is still in the port channel

183
00:13:03,000 --> 00:13:07,000
and from a Spanning Tree point of view

184
00:13:07,000 --> 00:13:10,000
the port channel is still up and still forwarding.

185
00:13:10,000 --> 00:13:13,000
The difference here, however, is the cost has increased

186
00:13:13,000 --> 00:13:16,000
because one of the ports has gone down.

187
00:13:16,000 --> 00:13:21,000
go back into the interface and no shut it.

188
00:13:21,000 --> 00:13:24,000
Look at Spanning Tree again

189
00:13:24,000 --> 00:13:31,000
port is coming up, the root port is still port 0/1

190
00:13:31,000 --> 00:13:37,000
looking at the summary notice both ports are part of the link aggregation now.

191
00:13:37,000 --> 00:13:42,000
So as we can see the port channel is still used by Spanning Tree

192
00:13:42,000 --> 00:13:45,000
it may take it a while for everything to converge

193
00:13:45,000 --> 00:13:50,000
if we'll look at the port channel details.

194
00:13:50,000 --> 00:14:00,000
Notice both ports are now active in the port channel a minute and 26 seconds ago

195
00:14:00,000 --> 00:14:04,000
gigabit 0/0 was unbundled from the port channel

196
00:14:04,000 --> 00:14:11,000
and then 47 seconds ago it was bundled back into the link aggregation or EtherChannel.

197
00:14:11,000 --> 00:14:17,000
So that’s an example of setting up a layer 2 link aggregation between 2 switches

198
00:14:17,000 --> 00:14:20,000
which provides for additional redundancy

199
00:14:20,000 --> 00:14:25,000
additional throughput and stops Spanning Tree blocking ports unnecessarily.
