1
00:00:00,940 --> 00:00:02,750
Okay, haven't found any VMs.

2
00:00:02,750 --> 00:00:05,290
I'm not exactly sure what that is,

3
00:00:05,290 --> 00:00:10,840
maybe my on‑premises VMware machines haven't been online all that long,

4
00:00:10,840 --> 00:00:14,750
but this is where you can enumerate all of the virtual machines that have

5
00:00:14,750 --> 00:00:18,710
been detected on your local or off‑cloud environment,

6
00:00:18,710 --> 00:00:19,240
okay?

7
00:00:19,240 --> 00:00:23,070
And then you bring in again replication metadata and then let it rip.

8
00:00:23,070 --> 00:00:25,610
I know I'm being relatively schematic.

9
00:00:25,610 --> 00:00:30,360
I'm doing that intentionally because the Azure architecture exams get into

10
00:00:30,360 --> 00:00:34,040
the weeds much more so, so does the Azure Administrator,

11
00:00:34,040 --> 00:00:36,080
then does the AZ‑801.

12
00:00:36,080 --> 00:00:39,550
For AZ‑801, you need to know the basic workflow.

13
00:00:39,550 --> 00:00:41,360
That when you're doing Hyper‑V,

14
00:00:41,360 --> 00:00:45,090
you just install that recovery provider on your Hyper‑V hosts.

15
00:00:45,090 --> 00:00:47,090
You don't need a separate appliance.

16
00:00:47,090 --> 00:00:49,600
If you're protecting on‑premises VMware,

17
00:00:49,600 --> 00:00:52,240
you'll need them orchestrated with vCenter,

18
00:00:52,240 --> 00:00:55,730
and you will have a local virtual machine appliance that

19
00:00:55,730 --> 00:00:57,940
helps you with inventory and so on.

20
00:00:57,940 --> 00:01:01,960
And you're specifying those machines that comprise your workload and

21
00:01:01,960 --> 00:01:04,290
where the recovery environment is going to be.

22
00:01:04,290 --> 00:01:08,140
I'm assuming that we're all going to failover into Azure.

23
00:01:08,140 --> 00:01:12,440
Now, if your machines are already in Azure arranged in a workload,

24
00:01:12,440 --> 00:01:14,770
your prep step's already finished,

25
00:01:14,770 --> 00:01:19,230
so you can go right to enabling replication by specifying your source,

26
00:01:19,230 --> 00:01:21,460
what VMs you want to protect,

27
00:01:21,460 --> 00:01:25,030
and then the replication settings as you can see here,

28
00:01:25,030 --> 00:01:25,770
all right?

29
00:01:25,770 --> 00:01:27,130
Now, as you can imagine,

30
00:01:27,130 --> 00:01:31,650
it's going to take the longest amount of time for that initial replication to

31
00:01:31,650 --> 00:01:35,450
happen between your source and disaster recovery environments.

32
00:01:35,450 --> 00:01:40,140
Let me head on over into my westus recovery services vault,

33
00:01:40,140 --> 00:01:42,920
and we'll come down under Protected Items,

34
00:01:42,920 --> 00:01:46,540
Replicated Items, and see what we have here.

35
00:01:46,540 --> 00:01:51,180
It looks like I have just a single virtual machine that is in a

36
00:01:51,180 --> 00:01:54,150
healthy replication state, and it is protected,

37
00:01:54,150 --> 00:01:59,890
which means that its storage and configuration are synchronized with this vault.

38
00:01:59,890 --> 00:02:03,240
That's going to be my disaster recovery environment here.

39
00:02:03,240 --> 00:02:07,240
So if we give it a click, we can come into its details here.

40
00:02:07,240 --> 00:02:11,500
And the idea is that you can come in and inspect each VM,

41
00:02:11,500 --> 00:02:17,050
but unless you specifically want to failover just a single virtual machine,

42
00:02:17,050 --> 00:02:19,090
this isn't the interface for it.

43
00:02:19,090 --> 00:02:22,720
Please understand, it took me a little while when I was first learning this.

44
00:02:22,720 --> 00:02:27,650
If you have a workload, you're going to want to failover the recovery plan,

45
00:02:27,650 --> 00:02:29,600
not the individual machines.

46
00:02:29,600 --> 00:02:30,640
Let me show you what I mean.

47
00:02:30,640 --> 00:02:32,440
Let me close out of here.

48
00:02:32,440 --> 00:02:36,420
And if we come down under Manage, Recovery Plans,

49
00:02:36,420 --> 00:02:39,540
let's create a new recovery plan right now.

50
00:02:39,540 --> 00:02:43,040
I'll call this dr‑plan1.

51
00:02:43,040 --> 00:02:47,940
The source is in East US, target is in West US,

52
00:02:47,940 --> 00:02:52,080
Resource Manager deployment, and we can select the items here,

53
00:02:52,080 --> 00:02:53,080
there's my web2.

54
00:02:53,080 --> 00:02:56,330
But again, the idea is this would be your collection.

55
00:02:56,330 --> 00:02:59,370
Maybe you have three web servers, three business servers,

56
00:02:59,370 --> 00:03:03,550
and six database servers, you can do a multi select here.

57
00:03:03,550 --> 00:03:07,940
I just have a simple, single VM environment, nothing too complicated.

58
00:03:07,940 --> 00:03:12,100
And also the abstraction of Azure Resource Manager means that this interface

59
00:03:12,100 --> 00:03:16,140
is the same regardless of whether web2 is running on‑prem,

60
00:03:16,140 --> 00:03:20,600
in another cloud, or as a native Azure virtual machine.

61
00:03:20,600 --> 00:03:22,590
In some sense, it doesn't matter.

62
00:03:22,590 --> 00:03:27,840
Let's click OK, and then let's click Create to build that recovery plan.

63
00:03:27,840 --> 00:03:32,840
The recovery plan is going to give us an anchor point to conduct our failovers,

64
00:03:32,840 --> 00:03:35,960
and it's also going to give us the opportunity to customize.

65
00:03:35,960 --> 00:03:38,230
Let's click in here, and speaking of which,

66
00:03:38,230 --> 00:03:43,240
let's click Customize to customize the behavior of the failover.

67
00:03:43,240 --> 00:03:46,320
Notice that the first step is always, all groups shut down.

68
00:03:46,320 --> 00:03:49,740
The idea is that when you're conducting a failover,

69
00:03:49,740 --> 00:03:53,610
you want to make sure that all the source machines are powered down.

70
00:03:53,610 --> 00:03:56,920
Now if you have a regional failure or a data center outage,

71
00:03:56,920 --> 00:03:58,380
then they're down anyway,

72
00:03:58,380 --> 00:04:01,810
so this first phase isn't going to make that big of a difference.

73
00:04:01,810 --> 00:04:06,140
So all groups shut down, and there's just the one machine in it right now.

74
00:04:06,140 --> 00:04:10,460
Notice that when you open the ellipsis, you can choose pre and post actions,

75
00:04:10,460 --> 00:04:13,340
and you can do that for all of your stages.

76
00:04:13,340 --> 00:04:16,840
What that means, let me go to pre‑action here for a moment,

77
00:04:16,840 --> 00:04:21,220
is that you can insert either a script or you can just

78
00:04:21,220 --> 00:04:24,810
provide action instructions that your engineers will see in

79
00:04:24,810 --> 00:04:28,160
the UI when they initiate a failover, okay?

80
00:04:28,160 --> 00:04:31,370
More useful though it seems to me would be to use a script.

81
00:04:31,370 --> 00:04:35,110
The script could, for example, verify services are started,

82
00:04:35,110 --> 00:04:37,130
verify file system artifacts,

83
00:04:37,130 --> 00:04:41,940
it can just serve as a nice check or a fix‑up step either before

84
00:04:41,940 --> 00:04:45,390
and/or after each phase during your failover.

85
00:04:45,390 --> 00:04:45,810
Now,

86
00:04:45,810 --> 00:04:50,910
the catch here is that you store your scripts in an Azure Automation account.

87
00:04:50,910 --> 00:04:53,130
You can't just upload a script ad hoc,

88
00:04:53,130 --> 00:04:57,540
it has to be a runbook that's registered in an Automation Account,

89
00:04:57,540 --> 00:04:58,780
so that's the idea.

90
00:04:58,780 --> 00:05:01,350
I don't have any here and sort of barking at me here,

91
00:05:01,350 --> 00:05:05,180
no runbooks found, yeah, it's empty, so it's not showing anything.

92
00:05:05,180 --> 00:05:07,490
But, you know, you have your auto account,

93
00:05:07,490 --> 00:05:10,100
you upload your scripts to it and specify it here.

94
00:05:10,100 --> 00:05:11,210
Not that big of a deal,

95
00:05:11,210 --> 00:05:15,360
but I just want you to see that that's how you would expand the capabilities of

96
00:05:15,360 --> 00:05:20,470
what's happening before and/or after each of these stages.

97
00:05:20,470 --> 00:05:24,850
Okay, so the next stage is all groups failover where,

98
00:05:24,850 --> 00:05:26,840
in this case, it's just the one machine.

99
00:05:26,840 --> 00:05:29,180
But notice that you can create groups,

100
00:05:29,180 --> 00:05:34,040
and you can change group identifications of selected virtual machines.

101
00:05:34,040 --> 00:05:37,390
The idea here is that you'll want to orchestrate all the VMs that

102
00:05:37,390 --> 00:05:40,380
are in each class of your workload into a group.

103
00:05:40,380 --> 00:05:43,010
You might have a database group, a business group,

104
00:05:43,010 --> 00:05:45,540
and a web front end group, for example.

105
00:05:45,540 --> 00:05:47,580
And then, very importantly,

106
00:05:47,580 --> 00:05:53,680
you can control the order in which those groups failover and then start up.

107
00:05:53,680 --> 00:05:54,940
You see what I mean?

108
00:05:54,940 --> 00:05:58,180
You might want your database tier up and running before your

109
00:05:58,180 --> 00:06:02,050
API tier and both of those tiers up and running before you

110
00:06:02,050 --> 00:06:04,340
bring up your web front end.

111
00:06:04,340 --> 00:06:08,720
Now I said this earlier in the lesson, but please remember that at this point,

112
00:06:08,720 --> 00:06:14,220
my web2 VM is not actually running in West US, it's just that the

113
00:06:14,220 --> 00:06:18,400
configuration is securely stored here in the recovery plan,

114
00:06:18,400 --> 00:06:20,780
and it's basically in a replicated state.

115
00:06:20,780 --> 00:06:23,340
It is in a replicated state, but it's turned off,

116
00:06:23,340 --> 00:06:26,000
very much like Hyper‑V Replica acts like,

117
00:06:26,000 --> 00:06:28,320
and you'll see that in the next module actually,

118
00:06:28,320 --> 00:06:29,440
okay?

119
00:06:29,440 --> 00:06:33,610
So these controls allow you to, if I select web2 here,

120
00:06:33,610 --> 00:06:35,760
we can remove, we can move.

121
00:06:35,760 --> 00:06:38,600
UI is a little bit tough to get used to,

122
00:06:38,600 --> 00:06:39,750
to be totally honest,

123
00:06:39,750 --> 00:06:45,560
but that's how you work with creating the actual plan for your failover.

124
00:06:45,560 --> 00:06:48,790
Now when it comes to failing over, you can do test or real.

125
00:06:48,790 --> 00:06:50,330
I'm going to jump right to real,

126
00:06:50,330 --> 00:06:53,540
so I can show you the business point of view here.

127
00:06:53,540 --> 00:06:55,960
Remember test does do a failover,

128
00:06:55,960 --> 00:07:00,890
but it allows you to clean up the built environment with a single mouse click.

129
00:07:00,890 --> 00:07:03,010
And it tells us here that it's recommended,

130
00:07:03,010 --> 00:07:08,050
are you sure you want to embrace this risk and skip Test Failover?

131
00:07:08,050 --> 00:07:09,240
I do.

132
00:07:09,240 --> 00:07:14,020
Now actually it looks like Azure is not allowing me to bypass that,

133
00:07:14,020 --> 00:07:19,980
so I guess Microsoft is really feeling strongly about us doing a Test Failover.

134
00:07:19,980 --> 00:07:22,940
So here we go, let's do a Test Failover.

135
00:07:22,940 --> 00:07:23,730
And once again,

136
00:07:23,730 --> 00:07:32,000
I want you to see that we're doing a failover of the recovery plan, not an individual machine.