WEBVTT

1
00:00.000 --> 00:02.900
So for our game plan for the NSX Federation,

2
00:02.900 --> 00:05.400
I already have implemented two sites.

3
00:05.400 --> 00:07.800
We'll call them site one and site two.

4
00:07.800 --> 00:08.900
They're all set up.

5
00:08.900 --> 00:11.400
They have local NSX managers already in place.

6
00:11.400 --> 00:15.100
They also have some tier zero and tier one gateways in place

7
00:15.100 --> 00:17.700
and also some local segments already in place.

8
00:17.700 --> 00:20.600
However, what's not in place is the global manager

9
00:20.600 --> 00:24.100
in any of the stretched or shared gateway or segments

10
00:24.100 --> 00:25.500
across those sites.

11
00:25.500 --> 00:27.500
So as far as the game plan is concerned,

12
00:27.500 --> 00:30.500
let's go ahead and do some prep work in our topologies
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

12
00:27.500 --> 00:30.500
let's go ahead and do some prep work in our topologies

13
00:30.500 --> 00:32.700
to make sure we are ready for success

14
00:32.700 --> 00:34.600
with the global NSX manager.

15
00:34.600 --> 00:36.500
And as part of that prep on each of the sites

16
00:36.500 --> 00:37.900
with their local managers,

17
00:37.900 --> 00:40.900
we need to set a default transport zone.

18
00:40.900 --> 00:43.600
And the reason we need that is because the global manager

19
00:43.600 --> 00:45.700
is going to look at for each of the sites,

20
00:45.700 --> 00:48.000
what is your default overlay transport zone

21
00:48.000 --> 00:51.000
and use that for any common stretch segments

22
00:51.000 --> 00:53.000
that are implemented by the global manager.

23
00:53.000 --> 00:54.900
So we want to make sure that site one

24
00:54.900 --> 00:56.800
and the NSX manager there and that site two,

25
00:56.800 --> 00:59.300
the NSX manager here that we have identified

26
00:59.300 --> 01:01.600
our default overlay transport zone.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

26
00:59.300 --> 01:01.600
our default overlay transport zone.

27
01:01.600 --> 01:04.900
We also want to make sure that for our NSX management cluster

28
01:04.900 --> 01:06.000
on each of the sites,

29
01:06.000 --> 01:09.900
we also want to make sure that we have set up a virtual IP address

30
01:09.900 --> 01:12.200
because that's what the global manager is going to be interacting

31
01:12.200 --> 01:13.200
and working with.

32
01:13.200 --> 01:16.200
And that's a good thing because if we lose one of our NSX managers

33
01:16.200 --> 01:17.000
in the cluster,

34
01:17.000 --> 01:20.400
we want the global manager to be able to continue working with that cluster.

35
01:20.400 --> 01:23.900
So we're going to have the global manager point to the virtual IP address

36
01:23.900 --> 01:27.500
instead of the actual address on any of the individual VMs

37
01:27.500 --> 01:29.000
that are in the management cluster.

38
01:29.000 --> 01:32.950
And we're also going to want to have a backup of our NSX management for site
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

38
01:29.000 --> 01:32.950
And we're also going to want to have a backup of our NSX management for site

39
01:32.950 --> 01:33.300
one

40
01:33.300 --> 01:34.200
and site two.

41
01:34.200 --> 01:38.500
In fact, before we can onboard completely with the sites

42
01:38.500 --> 01:40.600
and the NSX managers in those sites,

43
01:40.600 --> 01:42.500
it's going to require a backup.

44
01:42.500 --> 01:43.900
So we're going to take a preemptive strike

45
01:43.900 --> 01:46.400
and we're going to back up the local managers first

46
01:46.400 --> 01:48.100
that way we have a backup ready to go.

47
01:48.100 --> 01:49.200
So once we've done the prep,

48
01:49.200 --> 01:52.300
we're then going to go ahead and deploy the global manager.

49
01:52.300 --> 01:55.000
I'll put GM there for short and I'll label it here as well.

50
01:55.000 --> 01:58.640
There's our global manager and that process for the initial deployment of that

51
01:58.640 --> 01:58.900
global

52
01:58.900 --> 02:01.900
manager is very much like it is for a local manager.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

52
01:58.900 --> 02:01.900
manager is very much like it is for a local manager.

53
02:01.900 --> 02:04.100
We're going to deploy from an OVF template,

54
02:04.100 --> 02:07.310
specify the compute resource and the big difference here is when we're

55
02:07.310 --> 02:07.900
deploying this

56
02:07.900 --> 02:09.000
NSX manager,

57
02:09.000 --> 02:13.500
we're going to choose the option of a global manager and not just an NSX

58
02:13.500 --> 02:14.000
manager.

59
02:14.000 --> 02:15.700
And that's what changes its role.

60
02:15.700 --> 02:18.000
So it knows it's going to be a global manager.

61
02:18.000 --> 02:19.100
So as part of our deployment,

62
02:19.100 --> 02:20.800
we'll specify it's going to be a global manager.

63
02:20.800 --> 02:23.500
We also have to specify that it's going to be active.

64
02:23.500 --> 02:24.700
Now in a production environment,

65
02:24.700 --> 02:27.000
very similar to what we want to do with our local managers,

66
02:27.000 --> 02:28.200
we want to have a cluster,

67
02:28.200 --> 02:28.400
you know,

68
02:28.400 --> 02:31.900
we don't have at least three inside that cluster for fault tolerance.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

68
02:28.400 --> 02:31.900
we don't have at least three inside that cluster for fault tolerance.

69
02:31.900 --> 02:34.000
Same thing to true here for the global manager.

70
02:34.000 --> 02:35.900
So as part of that on this global manager,

71
02:35.900 --> 02:36.600
it's going to be asking,

72
02:36.600 --> 02:36.800
Hey,

73
02:36.800 --> 02:38.300
what is the compute manager?

74
02:38.300 --> 02:42.250
And the reason it needs a compute manager is for the deployment of the second

75
02:42.250 --> 02:43.400
and third

76
02:43.400 --> 02:46.600
NSX managers as part of that cluster.

77
02:46.600 --> 02:47.600
So my lab environment here,

78
02:47.600 --> 02:49.800
I'm just going to use a single global manager.

79
02:49.800 --> 02:50.400
However, again,

80
02:50.400 --> 02:53.200
if we wanted to create additional global managers,

81
02:53.200 --> 02:56.400
that's the reason the global manager wants to know about the compute resource

82
02:56.400 --> 02:58.000
for the second and third.

83
02:58.000 --> 02:59.600
So once that global manager is deployed,

84
02:59.600 --> 03:01.600
we're then going to go ahead and add sites.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

84
02:59.600 --> 03:01.600
we're then going to go ahead and add sites.

85
03:01.600 --> 03:02.600
So in our topology,

86
03:02.600 --> 03:05.200
we're going to add site one and site two.

87
03:05.200 --> 03:06.400
And when we onboard these sites,

88
03:06.400 --> 03:09.410
we're going to specify the virtual IP address for the management cluster at

89
03:09.410 --> 03:10.200
site one

90
03:10.200 --> 03:12.600
and put in the details for the authentication there

91
03:12.600 --> 03:14.600
and a similar process for site two,

92
03:14.600 --> 03:16.400
including specifying the IP address

93
03:16.400 --> 03:20.120
or a fully qualified domain name that points to the virtual IP address for the

94
03:20.120 --> 03:21.600
management cluster at site two,

95
03:21.600 --> 03:24.200
put in the credentials and onboard that as well.

96
03:24.200 --> 03:25.500
And then once we've added both sites,

97
03:25.500 --> 03:29.200
we're then going to import the configurations from both sites.

98
03:29.200 --> 03:29.600
Also,

99
03:29.600 --> 03:34.800
just be aware that not every single feature is supported at the global manager.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

99
03:29.600 --> 03:34.800
just be aware that not every single feature is supported at the global manager.

100
03:34.800 --> 03:36.600
So before you deploy this in production,

101
03:36.600 --> 03:39.400
you want to take a look and see what features are supported.

102
03:39.400 --> 03:42.260
And if there's some critical function that you need that's not supported by the

103
03:42.260 --> 03:43.200
global manager,

104
03:43.200 --> 03:45.200
you'd want to be aware of that before rolling it out.

105
03:45.200 --> 03:48.560
So at this point, the global manager would have imported the configs from both

106
03:48.560 --> 03:49.000
sites.

107
03:49.000 --> 03:52.500
So our next step would be to go ahead and configure the remote tunnel endpoint

108
03:52.500 --> 03:54.600
addresses for our edges.

109
03:54.600 --> 03:55.200
So the edges,

110
03:55.200 --> 03:58.800
they can still use the same uplink that they're using for their taps.

111
03:58.800 --> 04:01.500
However, we want to use a separate VLAN.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

111
03:58.800 --> 04:01.500
However, we want to use a separate VLAN.

112
04:01.500 --> 04:04.700
So for the two edges here at site one for edge one,

113
04:04.700 --> 04:11.480
let's use VLAN 71 and for edge two, let's use VLAN 72 and over here at site two

114
04:11.480 --> 04:14.600
for the existing two edges that are currently there for edge one.

115
04:14.600 --> 04:19.600
Let's use VLAN 81 and for edge two, we'll use VLAN 82.

116
04:19.600 --> 04:24.280
And as far as IP addressing goes for the VLAN 71, I'm going to use 71 71, et c

117
04:24.280 --> 04:24.800
etera.

118
04:24.800 --> 04:29.430
And for VLAN 72, will you 72 72 for the IP addressing for that VLAN and a

119
04:29.430 --> 04:32.100
similar treatment for VLAN 81 and 82 that way.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

119
04:29.430 --> 04:32.100
similar treatment for VLAN 81 and 82 that way.

120
04:32.100 --> 04:35.730
Once it's configured, we can take a look at the IP addresses involved for those

121
04:35.730 --> 04:37.700
remote taps and know exactly what they are.

122
04:37.700 --> 04:41.100
Now, once we have that all set up, it'd be wise to go ahead and test it.

123
04:41.100 --> 04:46.000
So to test it, let's create one single gateway, stretch gateway and let's go

124
04:46.000 --> 04:50.600
ahead and create a tier one gateway and a shared segment that we can go ahead

125
04:50.600 --> 04:53.100
and implement across both of our sites.

126
04:53.100 --> 04:56.520
So this will represent our stretch segment right here and let's go ahead and

127
04:56.520 --> 04:58.300
call that segment segment 50.

128
04:58.300 --> 05:02.390
So over here at site one, I've got segment 10 and 20 over here at site two, we
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

128
04:58.300 --> 05:02.390
So over here at site one, I've got segment 10 and 20 over here at site two, we

129
05:02.390 --> 05:03.900
've got segment 30 and 40.

130
05:03.900 --> 05:08.540
So we'll create a new stretch segment called segment 50 that'll be available to

131
05:08.540 --> 05:09.200
both sites.

132
05:09.200 --> 05:13.500
And then we can connect a VM here and here to that common segment.

133
05:13.500 --> 05:19.370
And for this VM here, let's use the IP address of 10.50.0.11 and over here on

134
05:19.370 --> 05:25.200
this VM over at site two, let's use the address of 10.50.0.22.

135
05:25.200 --> 05:29.620
So the traffic between those two VMs would go through their respective host

136
05:29.620 --> 05:33.730
transport nodes who had forwarded to the local edges, who would then tunnel it
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

136
05:29.620 --> 05:33.730
transport nodes who had forwarded to the local edges, who would then tunnel it

137
05:33.730 --> 05:37.300
over to the other site respectively using the R-type addresses.

138
05:37.300 --> 05:40.520
And then just for good measure, let's also demonstrate how we can use the

139
05:40.520 --> 05:44.820
distributed firewall and implement some rules using the global manager to

140
05:44.820 --> 05:48.300
control the traffic and what's allowed to go between the two sites.

141
05:48.300 --> 05:52.890
So with that game plan in mind, join me in the next video as we do our initial

142
05:52.890 --> 05:56.300
prep work on the local managers here at site one and site two.