WEBVTT

1
00:00.210 --> 00:02.370
If we wanna implement routing,

2
00:02.370 --> 00:04.740
for example, having these three routers

3
00:04.740 --> 00:06.450
to do logical routing of our traffic

4
00:06.450 --> 00:09.480
for north-south traffic and east-west traffic.

5
00:09.480 --> 00:10.740
To implement that functionality,

6
00:10.740 --> 00:13.380
we're gonna configure it all from the NSX Manager,

7
00:13.380 --> 00:15.390
so you and I interface with the NSX Manager

8
00:15.390 --> 00:17.970
and then we specify that we want to have, for example,

9
00:17.970 --> 00:20.580
gateway zero, and gateway A, and gateway B,

10
00:20.580 --> 00:24.480
and connect them logically to those various NSX segments.

11
00:24.480 --> 00:29.250
And when we create this logical routing inside of NSX,

12
00:29.250 --> 00:32.190
the actual ESXi host transport nodes,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

12
00:29.250 --> 00:32.190
the actual ESXi host transport nodes,

13
00:32.190 --> 00:35.580
they can actually do the logical routing themselves,

14
00:35.580 --> 00:36.840
and then if they need to forward traffic

15
00:36.840 --> 00:40.110
to another ESXi host, they can go ahead and tunnel it over.

16
00:40.110 --> 00:42.360
And that's known as a distributed router,

17
00:42.360 --> 00:45.150
where the logical routers are implemented

18
00:45.150 --> 00:47.700
on the actual host transport nodes.

19
00:47.700 --> 00:48.600
However, sometimes

20
00:48.600 --> 00:50.820
we may need to integrate additional services,

21
00:50.820 --> 00:54.510
including things like network address translation, or VPNs,

22
00:54.510 --> 00:57.180
or connecting our NSX routed environment

23
00:57.180 --> 01:00.900
out to the outside world, and for those specific functions
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

23
00:57.180 --> 01:00.900
out to the outside world, and for those specific functions

24
01:00.900 --> 01:03.780
that can't be done by the host transport nodes themselves,

25
01:03.780 --> 01:05.610
we have a separate entity,

26
01:05.610 --> 01:09.240
and that, my friend, is an edge transport node.

27
01:09.240 --> 01:10.500
Now we have a couple of options

28
01:10.500 --> 01:12.450
for implementing an edge transport node,

29
01:12.450 --> 01:14.340
the easiest option in my opinion,

30
01:14.340 --> 01:16.380
is to implement a little VM.

31
01:16.380 --> 01:18.630
So this little VM is running on a host somewhere,

32
01:18.630 --> 01:22.290
and this little VM is implemented, and it's an edge node.

33
01:22.290 --> 01:23.910
So we can have an edge node VM,

34
01:23.910 --> 01:26.910
or we could also do a bare-metal edge node.

35
01:26.910 --> 01:29.730
And that's where, instead of running the edge node as a VM,

36
01:29.730 --> 01:31.440
we'd take a physical computer
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

36
01:29.730 --> 01:31.440
we'd take a physical computer

37
01:31.440 --> 01:33.180
and install the appropriate software there.

38
01:33.180 --> 01:35.430
And so one of the questions might be why would we ever,

39
01:35.430 --> 01:37.380
you know, wanna have a dedicated piece of hardware

40
01:37.380 --> 01:39.270
as an edge node when we can implement

41
01:39.270 --> 01:41.670
the functionality of these gateways as a VM?

42
01:41.670 --> 01:43.350
And the answer is throughput.

43
01:43.350 --> 01:44.400
If you wanna crank out

44
01:44.400 --> 01:47.340
20, or 40, or 100 gigabits per second,

45
01:47.340 --> 01:49.650
you may need some dedicated hardware to do that,

46
01:49.650 --> 01:52.920
and that's why there's the option of a bare-metal edge node.

47
01:52.920 --> 01:54.420
So let's imagine we wanna implement

48
01:54.420 --> 01:58.050
the routing infrastructure and we decide to deploy a VM.

49
01:58.050 --> 02:00.750
So to deploy that VM for the edge node,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

49
01:58.050 --> 02:00.750
So to deploy that VM for the edge node,

50
02:00.750 --> 02:03.840
as I mentioned, it's done right through the NSX Manager

51
02:03.840 --> 02:07.530
and you simply specify what host you wanna run that VM on,

52
02:07.530 --> 02:09.150
we're gonna specify which uplinks

53
02:09.150 --> 02:11.940
on that host should be used by the edge node,

54
02:11.940 --> 02:15.150
and let's go ahead and visualize that with an ESXi host.

55
02:15.150 --> 02:17.130
Let's say this is an ESXi host here,

56
02:17.130 --> 02:20.640
and here's our edge node VM running on that host,

57
02:20.640 --> 02:23.100
and of course the host has some physical connectivity out,

58
02:23.100 --> 02:25.080
and as part of our configuration of the edge node,

59
02:25.080 --> 02:27.120
we're also gonna specify which uplinks

60
02:27.120 --> 02:28.530
we want the edge node to use.

61
02:28.530 --> 02:30.480
So let's say I have VM NIC zero, and one, and two,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

61
02:28.530 --> 02:30.480
So let's say I have VM NIC zero, and one, and two,

62
02:30.480 --> 02:31.830
and three, and four.

63
02:31.830 --> 02:34.140
And very similar to the uplink profile

64
02:34.140 --> 02:36.000
that we specified for an ESXi host

65
02:36.000 --> 02:38.460
when it's acting as a host transport node,

66
02:38.460 --> 02:40.530
we're also, when we deploy the edge node,

67
02:40.530 --> 02:43.710
gonna specify which uplinks we want it to use

68
02:43.710 --> 02:45.480
on the host that it's currently running on.

69
02:45.480 --> 02:47.130
So, maybe we want this edge node to go ahead

70
02:47.130 --> 02:50.160
and use VM NIC three as the active

71
02:50.160 --> 02:52.230
and then VM NIC four as the standby,

72
02:52.230 --> 02:54.090
I'll put a A there and S there.

73
02:54.090 --> 02:55.560
And as a result, that edge node

74
02:55.560 --> 02:57.930
is also going to get a tunnel endpoint.

75
02:57.930 --> 03:00.210
So that tunnel endpoint, we'll go ahead and call this one,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

75
02:57.930 --> 03:00.210
So that tunnel endpoint, we'll go ahead and call this one,

76
03:00.210 --> 03:04.860
let's say it's 10.33.33.33.

77
03:04.860 --> 03:06.000
And another question might be, well,

78
03:06.000 --> 03:07.380
what would this need to tunnel to?

79
03:07.380 --> 03:08.213
And the answer is

80
03:08.213 --> 03:10.890
it could tunnel to any other transport node,

81
03:10.890 --> 03:14.220
so we could tunnel over here to ESXi, let's say ESXi-A,

82
03:14.220 --> 03:16.410
which is acting as a host transport node

83
03:16.410 --> 03:20.010
with this tunnel endpoint address of 10.11.11.11.

84
03:20.010 --> 03:23.850
We can add this edge node with a tunnel going to ESXi-B,

85
03:23.850 --> 03:28.733
and perhaps this tunnel endpoint address is 10.22.22.22.

86
03:29.670 --> 03:30.630
And the cool thing is
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

86
03:29.670 --> 03:30.630
And the cool thing is

87
03:30.630 --> 03:32.820
that there is a function called a distributed router,

88
03:32.820 --> 03:33.960
we'll call it a DR for short,

89
03:33.960 --> 03:36.630
not to be confused with an OSPF designated router,

90
03:36.630 --> 03:39.180
so I'll spell it out, that's a distributed router.

91
03:39.180 --> 03:41.220
And that distributed router exists here

92
03:41.220 --> 03:43.620
on the edge node on the local control plane,

93
03:43.620 --> 03:47.880
it also exists here on ESXi-B and the ESXi-A,

94
03:47.880 --> 03:49.740
and the entity that's coordinating all that

95
03:49.740 --> 03:51.840
is the NSX Manager,

96
03:51.840 --> 03:53.760
and up here we have the centralized control plane.

97
03:53.760 --> 03:55.920
So the benefit of having a centralized control plane

98
03:55.920 --> 03:58.590
and then a local control plane that's been coordinated

99
03:58.590 --> 04:00.960
between the NSX Manager and that cluster,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

99
03:58.590 --> 04:00.960
between the NSX Manager and that cluster,

100
04:00.960 --> 04:04.410
and each of our nodes, which includes the ESXi host,

101
04:04.410 --> 04:07.410
and also now the edge node implemented as a VM,

102
04:07.410 --> 04:10.380
the benefit is if we have a VM list, call this VM-C,

103
04:10.380 --> 04:12.750
and let's say that VM is over here on segment 30,

104
04:12.750 --> 04:16.020
and that VM is sending traffic into the network,

105
04:16.020 --> 04:18.960
and that VM is connected to this logical segment 30,

106
04:18.960 --> 04:21.690
ESXi-B knows exactly how to forward it.

107
04:21.690 --> 04:23.850
Is it gonna make the forwarding decision locally

108
04:23.850 --> 04:25.830
because it has all the information that needs to do that

109
04:25.830 --> 04:28.860
or is the destination reachable through one of our routers,

110
04:28.860 --> 04:30.600
AKA an NSX gateway?
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

110
04:28.860 --> 04:30.600
AKA an NSX gateway?

111
04:30.600 --> 04:33.000
And if it is, ESXi-B can take it,

112
04:33.000 --> 04:34.710
encapsulate it into the tunnel,

113
04:34.710 --> 04:36.570
send it to the other side where it needs to go,

114
04:36.570 --> 04:38.580
where it can be deencapsulated and processed.

115
04:38.580 --> 04:41.340
Most of the time we're not just gonna have one edge node,

116
04:41.340 --> 04:43.020
we'll have an edge node cluster.

117
04:43.020 --> 04:46.080
So let's say this is ESXi-10,

118
04:46.080 --> 04:49.459
and let's also draw a really thin ESXi host over here,

119
04:49.459 --> 04:50.730
(laughs) because of the space, so.

120
04:50.730 --> 04:53.640
So, here's ESXi-11,

121
04:53.640 --> 04:56.610
and it also has an edge node VM,

122
04:56.610 --> 04:59.580
and we can implement those as a cluster that work together.

123
04:59.580 --> 05:01.710
Now here's the great news, in this skill,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

123
04:59.580 --> 05:01.710
Now here's the great news, in this skill,

124
05:01.710 --> 05:06.710
I wanted to present a big picture look at the concept of NSX

125
05:06.720 --> 05:09.540
and the key players inside that architecture.

126
05:09.540 --> 05:12.390
So with this warmup to the world of NSX,

127
05:12.390 --> 05:13.710
what we're gonna do is we'll take a look

128
05:13.710 --> 05:16.410
at implementing and verifying each of the components

129
05:16.410 --> 05:19.083
in one nice little step at a time.