0
1
00:00:00,900 --> 00:00:03,600
In the middle we've got Joe from Docker.
1

2
00:00:03,600 --> 00:00:08,580
Joe is the director of engineering and he's going to be sharing some really cool new features in the
2

3
00:00:08,570 --> 00:00:12,120
19.03 release of our Desktop Enterprise.
3

4
00:00:12,120 --> 00:00:17,670
So first up we've got Joe he's actually Where are you from today.
4

5
00:00:17,670 --> 00:00:21,120
I forget Indianapolis Indiana Indianapolis.
5

6
00:00:21,260 --> 00:00:21,570
Yes.
6

7
00:00:21,580 --> 00:00:23,710
So he's in the duck the office there right.
7

8
00:00:24,220 --> 00:00:24,750
Yeah that's right.
8

9
00:00:25,830 --> 00:00:30,720
And Joe is gonna talk at best a little bit about clustering.
9

10
00:00:30,730 --> 00:00:34,200
This is a brand new thing in Docker desktop.
10

11
00:00:34,200 --> 00:00:40,140
Well we've never really had a command line interface to create server clusters.
11

12
00:00:40,140 --> 00:00:44,580
And you know before we basically had little tools like Docker machine which would spend up a single
12

13
00:00:44,580 --> 00:00:48,160
instance but this is this is nothing like that right Joe.
13

14
00:00:48,460 --> 00:00:49,670
Right of course.
14

15
00:00:50,010 --> 00:00:56,700
It goes for the next level of additional functionality to sort of take care of the entire lifecycle
15

16
00:00:56,700 --> 00:01:00,300
management of a cluster like we do with containers and services.
16

17
00:01:00,300 --> 00:01:00,560
Yeah.
17

18
00:01:00,570 --> 00:01:06,510
And this from my understanding the history of this was around the DCI that I think I talked a lot about
18

19
00:01:06,510 --> 00:01:09,060
last year the docker certified infrastructure.
19

20
00:01:09,060 --> 00:01:11,020
And this is sort of the evolution of that is that right.
20

21
00:01:11,460 --> 00:01:12,010
Yes.
21

22
00:01:12,160 --> 00:01:12,380
Right.
22

23
00:01:12,390 --> 00:01:15,420
So the first sort of DCI templates and scripts.
23

24
00:01:15,430 --> 00:01:21,180
We we published a lot of good feedback from those and it was largely around.
24

25
00:01:21,180 --> 00:01:21,510
OK.
25

26
00:01:21,510 --> 00:01:26,250
Well I don't want to learn like a whole nother set of tooling and such.
26

27
00:01:26,250 --> 00:01:29,560
I mean thanks for the scripts and templates that would be useful.
27

28
00:01:29,610 --> 00:01:38,040
So we did an internal sort of revision three years to three where we put it all into a container and
28

29
00:01:38,040 --> 00:01:42,050
then we provided just a very simple command line interface to it.
29

30
00:01:42,120 --> 00:01:50,240
And so in this iteration which we call Docker cluster we're providing a declarative syntax to basically
30

31
00:01:50,240 --> 00:01:51,740
model the entire clusters.
31

32
00:01:52,120 --> 00:01:59,510
So you can add subtract nodes load balancers and such.
32

33
00:01:59,510 --> 00:01:59,910
All right.
33

34
00:01:59,920 --> 00:02:04,360
So it really cleans up the entire experience.
34

35
00:02:04,400 --> 00:02:05,500
Great.
35

36
00:02:05,540 --> 00:02:10,530
I know that a lot of us you know once we get past that darker one to one and we learn about building
36

37
00:02:10,640 --> 00:02:15,560
our containers and deploying our images to registries that's one of the next big questions is where
37

38
00:02:15,560 --> 00:02:16,070
do I put it.
38

39
00:02:16,910 --> 00:02:23,360
And I think that along the learning journey as well as those that are just more sysadmin ops like myself
39

40
00:02:23,390 --> 00:02:28,310
where they're focused more on the infrastructure side trying to figure out how to why you know how do
40

41
00:02:28,310 --> 00:02:33,590
I build out clusters of it of advanced solutions that aren't just sort of a basic machine with Docker
41

42
00:02:33,590 --> 00:02:34,390
running right.
42

43
00:02:34,520 --> 00:02:39,380
And I guess that's what this is really focused around right this is more about the enterprise clusters
43

44
00:02:39,380 --> 00:02:43,070
and not so much a one off machine with Docker running on it.
44

45
00:02:43,070 --> 00:02:43,880
Yeah exactly.
45

46
00:02:43,880 --> 00:02:53,160
So Docker machine is the current one offer for Docker running that I would say you should use.
46

47
00:02:53,240 --> 00:02:54,850
We have another tool that we do a lot of.
47

48
00:02:54,860 --> 00:03:02,360
Dr. Khan called Docker job which also uses a lot of these same technology stacks that I've been describing
48

49
00:03:03,350 --> 00:03:09,440
that once we're still working on improving that I think we'll have some pretty exciting announcements
49

50
00:03:09,440 --> 00:03:10,670
in about six months or so.
50

51
00:03:10,970 --> 00:03:11,670
Mm hmm.
51

52
00:03:11,720 --> 00:03:12,560
Cool.
52

53
00:03:12,560 --> 00:03:12,750
Yeah.
53

54
00:03:12,830 --> 00:03:16,700
Here first six months you put it on the internet is forever.
54

55
00:03:16,710 --> 00:03:18,720
So now now we're gonna be back here in six months.
55

56
00:03:20,090 --> 00:03:26,240
Well Joe so I know you have some demos to show and of course you know this everybody is watching so
56

57
00:03:26,240 --> 00:03:27,320
let's let's dive in.
57

58
00:03:27,320 --> 00:03:29,000
What what you got to show us.
58

59
00:03:29,000 --> 00:03:30,290
Sure of course.
59

60
00:03:30,800 --> 00:03:33,110
We also see my screen and you click.
60

61
00:03:33,430 --> 00:03:35,560
Yeah let me click some buttons
61

62
00:03:39,220 --> 00:03:44,070
so yeah you should be able to share your screen and Skype and then I should see that.
62

63
00:03:44,070 --> 00:03:44,320
All right
63

64
00:03:48,360 --> 00:03:51,200
green all right.
64

65
00:03:52,190 --> 00:03:59,930
OK so you know the place where most people start is by running dunker and with 19 3 we added the ability
65

66
00:03:59,930 --> 00:04:07,330
to put CLIA plug ins into Docker itself so you can add additional functionality.
66

67
00:04:07,370 --> 00:04:14,350
So we wrapped up all the functionality from DCI and put it in a tool called a cluster.
67

68
00:04:14,510 --> 00:04:22,460
So if we take a look at the functionality Docker cluster we have you know create create a cluster we
68

69
00:04:22,460 --> 00:04:29,390
can inspect the clusters we have we can list the ones we have some of the key features here.
69

70
00:04:29,390 --> 00:04:35,390
Being able to backup and restore clusters so that you know if something goes wrong you can always get
70

71
00:04:35,390 --> 00:04:37,210
back to a known good state.
71

72
00:04:38,330 --> 00:04:44,510
And then you know over the life cycle of the cluster being able to add nodes removed nodes upgrade from
72

73
00:04:44,840 --> 00:04:50,550
1983 to 1989 or whatever the next iteration will be.
73

74
00:04:50,550 --> 00:04:58,700
And that includes all of our offering as well as DCP DDR just making that whole upgrade experience seamless
74

75
00:04:58,700 --> 00:04:59,360
as possible.
75

76
00:05:00,860 --> 00:05:06,590
So what does one of these declarations look like.
76

77
00:05:06,680 --> 00:05:10,970
Take a look at the script that I have compared to here.
77

78
00:05:11,810 --> 00:05:17,030
So we have you know sort of a high level concept of variables where you can assign values so you can
78

79
00:05:17,030 --> 00:05:21,420
make templates for your for your clusters.
79

80
00:05:21,590 --> 00:05:27,110
And then as you deploy them you can sort of fill out unique values passwords of course we want to keep
80

81
00:05:27,110 --> 00:05:27,580
secret.
81

82
00:05:27,590 --> 00:05:36,260
So we'll make that into our declaration basic configuration to talk to NWS what regionally waiting to
82

83
00:05:36,260 --> 00:05:39,380
deploy and let's add some tags.
83

84
00:05:39,380 --> 00:05:43,430
Doctor we have a tool called Reaper that kills off old clusters.
84

85
00:05:43,430 --> 00:05:48,440
So I have to let it know that this is a pet and it's under the DCI project.
85

86
00:05:48,450 --> 00:05:51,850
So please don't kill it reaper.
86

87
00:05:52,340 --> 00:05:58,150
So then in terms of the cluster we can specify the engine version we're running to provide you C.P.
87

88
00:05:59,560 --> 00:06:02,780
the version we're wanting to install there.
88

89
00:06:03,110 --> 00:06:09,070
In this example just for brevity I'm only going to be installing CPD.
89

90
00:06:09,110 --> 00:06:11,860
Will also could also be installed.
90

91
00:06:13,100 --> 00:06:16,610
But again we don't take all day watching the install go.
91

92
00:06:17,600 --> 00:06:20,780
So then the resource types we can do eight of those instances.
92

93
00:06:20,780 --> 00:06:24,970
We can do a W spotted six requests to save some cash.
93

94
00:06:25,040 --> 00:06:30,620
We can do eight of us load balancers and DNS provisioning directly in here.
94

95
00:06:31,070 --> 00:06:38,690
So we could put a you know if you wanted your apps dot domain dot com we could have that pre provisioned
95

96
00:06:38,690 --> 00:06:41,700
with certificates provided by Let's Encrypt.
96

97
00:06:41,810 --> 00:06:48,110
So really the whole thing of all those different technologies are coming together under the under the
97

98
00:06:48,110 --> 00:06:52,010
hood here with with Dr. cluster.
98

99
00:06:52,490 --> 00:06:56,290
So what you get is you'll get those resources.
99

100
00:06:56,300 --> 00:07:00,370
It will also create what's called the docker context for you.
100

101
00:07:00,680 --> 00:07:08,670
So you can then switch your doctor client to that context and be able to interact with the cluster immediately.
101

102
00:07:09,290 --> 00:07:15,210
So we'll take a break from looking at that and I'll fire up
102

103
00:07:18,080 --> 00:07:31,210
the screen Dr. cluster eight minutes as Tucker and Deb show and call it plus we'll cooperate.
103

104
00:07:33,230 --> 00:07:37,640
So we need to provide our UCB password entering it and securely.
104

105
00:07:38,150 --> 00:07:39,490
So there we go.
105

106
00:07:39,490 --> 00:07:43,510
We'll talk to hub to look up licenses that I have available.
106

107
00:07:43,610 --> 00:07:50,090
We'll start planning the cluster on NWS and then we'll sort of see the slip are going from left to right
107

108
00:07:50,120 --> 00:07:58,190
letting us know that resources are permitting and then cooking show style.
108

109
00:07:58,190 --> 00:08:00,590
We'll take a look at the clusters we have available.
109

110
00:08:00,590 --> 00:08:04,430
And I've already set up a demo one in the background so we can go ahead and take a look at that right
110

111
00:08:04,460 --> 00:08:07,380
now.
111

112
00:08:10,630 --> 00:08:12,370
Doo doo doo doo doo
112

113
00:08:17,930 --> 00:08:27,780
and so there we can see this is when we expect the cluster we basically get a reflection of of the cluster
113

114
00:08:28,230 --> 00:08:33,480
declaration that we've provided to create with in addition to all the different
114

115
00:08:36,030 --> 00:08:43,710
parameters that were had default values they got resolved throughout the operation of the creation.
115

116
00:08:44,280 --> 00:08:50,940
So for instance there are some Cuba and options that got resolved because their defaults.
116

117
00:08:51,060 --> 00:08:59,280
The manager got an instance instantiated for it the instance type was this default instance type that
117

118
00:08:59,280 --> 00:09:02,740
we pick as well as we had choose a..
118

119
00:09:02,760 --> 00:09:08,980
So we chose it into a tool for as the default you can of course set it to any the supported OS as we
119

120
00:09:09,270 --> 00:09:12,270
have on our compatibility matrix or we'll pick a
120

121
00:09:15,110 --> 00:09:23,280
EMI that that works and that's that's known to be supported that's available and then we install everything
121

122
00:09:23,280 --> 00:09:24,820
else on top of that for you.
122

123
00:09:25,340 --> 00:09:35,640
So if you need to bring your own EMI to the cluster we can override this and say the explicit EMI that
123

124
00:09:35,640 --> 00:09:37,500
you need.
124

125
00:09:37,500 --> 00:09:38,810
So that's great.
125

126
00:09:38,880 --> 00:09:40,560
But how do we actually access it.
126

127
00:09:40,770 --> 00:09:44,170
So I mentioned that we have these things called the contexts.
127

128
00:09:44,190 --> 00:09:53,130
So right now I'm talking to the default context that I'm connected to locally on my doctor desktop.
128

129
00:09:53,760 --> 00:10:04,170
But what I want to do is I want to use this demo context which is the one for this cluster.
129

130
00:10:04,170 --> 00:10:06,780
So first just what the default context.
130

131
00:10:06,780 --> 00:10:10,110
We'll take a look and see that I have a couple of containers running.
131

132
00:10:10,110 --> 00:10:13,410
One is the one that's provisioning the cluster in the background.
132

133
00:10:14,640 --> 00:10:18,140
And another is a Vault Server I have running locally.
133

134
00:10:18,690 --> 00:10:26,900
But what we really want to do is take a look at the clusters the context you demo just like that or
134

135
00:10:26,910 --> 00:10:33,700
switch the demo and we can see that by doing a piece and taking a look and see all these wonderful you
135

136
00:10:33,720 --> 00:10:36,790
see containers running.
136

137
00:10:37,440 --> 00:10:50,790
We can also do Doctor info and see that the server is running 33 container let's go all that's supporting
137

138
00:10:50,790 --> 00:10:57,250
information in the cloud and where's all this stored.
138

139
00:10:57,250 --> 00:10:58,730
Is it just in the command line.
139

140
00:10:58,750 --> 00:10:59,810
The inspect.
140

141
00:10:59,830 --> 00:11:03,500
What am I inspecting when you say so.
141

142
00:11:03,710 --> 00:11:13,630
So in the doctor folder we have we have some metadata that we keep in in the background of the set of
142

143
00:11:13,630 --> 00:11:22,190
clusters you have the information for those clusters like your inventory SS age.
143

144
00:11:22,310 --> 00:11:28,630
Keys that we generate for you to just make it all is seamless and simple as possible.
144

145
00:11:28,660 --> 00:11:35,500
And additionally you can take this in and use the underlying tools that we use if you need to migrate
145

146
00:11:35,500 --> 00:11:39,930
to some of the right.
146

147
00:11:39,940 --> 00:11:42,540
Is this something like How would a team.
147

148
00:11:42,910 --> 00:11:48,190
How would a team of ops people use it if it's trying to think as if it's stored locally is there an
148

149
00:11:48,190 --> 00:11:51,410
option for like a remote store or something red.
149

150
00:11:51,430 --> 00:11:58,750
So to do that a remote state stored in this iteration there's not a where we're hard at work defining
150

151
00:11:58,750 --> 00:12:00,790
an API set up.
151

152
00:12:01,090 --> 00:12:06,910
That'll make it easy for someone to share the entire set functionality.
152

153
00:12:07,100 --> 00:12:10,960
Your organization team an individual level.
153

154
00:12:10,960 --> 00:12:12,900
Cool.
154

155
00:12:13,510 --> 00:12:16,130
And what versions of Docker are.
155

156
00:12:16,420 --> 00:12:19,280
Is this capable of deploying.
156

157
00:12:19,910 --> 00:12:23,830
So the ones that we support are Dr. Enterprise 3.0.
157

158
00:12:24,470 --> 00:12:29,590
That's the engine 1983 UCB 3 2 and Tier 2 7.
158

159
00:12:30,180 --> 00:12:30,890
OK.
159

160
00:12:30,980 --> 00:12:32,910
And I'm assuming you specify though I didn't.
160

161
00:12:32,940 --> 00:12:35,450
I missed it in the file but in the end you're specifying that.
161

162
00:12:35,450 --> 00:12:40,250
So like a future version you can you can define which versions.
162

163
00:12:40,830 --> 00:12:41,710
Yeah.
163

164
00:12:42,170 --> 00:12:44,880
So here's the 98 0 3.
164

165
00:12:45,500 --> 00:12:53,990
You can use some additional information here to drill down into a test channel if you're testing one
165

166
00:12:53,990 --> 00:13:03,750
of our betas and you can specify the like let's say you're putting your containers on an internal registry.
166

167
00:13:03,830 --> 00:13:14,210
You could modify this to have your own industry reference to to access.
167

168
00:13:14,240 --> 00:13:24,680
Are you there was a great question actually in chat about basically how would this be different than
168

169
00:13:24,680 --> 00:13:29,910
doing it in puppet or another you know like terraform or something like that is there.
169

170
00:13:31,310 --> 00:13:34,770
An advantage to this method.
170

171
00:13:34,860 --> 00:13:42,030
The primary advantage is that we've already done that work for you so you can just go ahead and start
171

172
00:13:42,150 --> 00:13:48,440
defining your cluster topology but underneath the hood you know very similar tools are being used.
172

173
00:13:48,450 --> 00:13:55,370
I think we're using terraform to create the resources so we're actually building out all those configurations
173

174
00:13:55,380 --> 00:14:02,550
behind the scenes where we built out a set of ethical rules and playbooks to drive that.
174

175
00:14:02,550 --> 00:14:10,230
So from that operate container you can copy out those things and integrate them into your system if
175

176
00:14:10,230 --> 00:14:10,870
necessary.
176

177
00:14:11,640 --> 00:14:13,840
OK very cool.
177

178
00:14:14,190 --> 00:14:16,710
And then you mentioned providers are there.
178

179
00:14:16,950 --> 00:14:19,620
What are the providers you have at lunch right.
179

180
00:14:19,620 --> 00:14:27,910
So at launch we launched with a W S I believe in this next patch release for nineteen thirty two.
180

181
00:14:28,410 --> 00:14:33,450
We'll be including the Azure supported hot on its heels.
181

182
00:14:33,450 --> 00:14:39,300
We'll be working on these sphere to be able to do this as well.
182

183
00:14:39,330 --> 00:14:40,920
All right.
183

184
00:14:42,570 --> 00:14:43,500
What else.
184

185
00:14:43,500 --> 00:14:46,880
So this is what Docker desktop enterprise right.
185

186
00:14:46,890 --> 00:14:48,540
This is enterprise.
186

187
00:14:48,840 --> 00:14:51,570
Well of course this is for deploying server enterprise.
187

188
00:14:51,610 --> 00:14:52,130
So.
188

189
00:14:52,690 --> 00:14:58,080
So it would make more sense for someone who is you know running a desktop enterprise so they get this
189

190
00:14:58,110 --> 00:14:59,730
and this is a command line plug in right.
190

191
00:14:59,730 --> 00:15:02,310
Which we kind of talked a little bit about plug ins yesterday.
191

192
00:15:02,310 --> 00:15:11,290
So if you're unfamiliar with plug ins the plug in model is one where the docker command line now Docker
192

193
00:15:11,420 --> 00:15:14,640
engine is has something called plugins for a while but the command line now has plug ins so that you
193

194
00:15:14,640 --> 00:15:21,360
can add extra functionality and a sneak hint at tomorrow is that we will have some other people on that
194

195
00:15:21,360 --> 00:15:25,790
show you how to make your own command line plugins to do basically whatever you want.
195

196
00:15:25,790 --> 00:15:30,600
And so the docker command line and essentially writing apps or scripts at the command line that have
196

197
00:15:30,600 --> 00:15:36,330
access to the docker functionality without you having to manually create all that all those API calls
197

198
00:15:36,330 --> 00:15:36,900
and stuff like that.
198

199
00:15:36,900 --> 00:15:38,500
So it's pretty neat.
199

200
00:15:38,520 --> 00:15:45,690
So this is taking advantage of that it looks like and the plugins are bundled into the enterprise deliver.
200

201
00:15:45,720 --> 00:15:51,870
So if you are installing the enterprise engine on on a node you can then use that as sort of a jump
201

202
00:15:51,870 --> 00:15:55,510
box to then create the rest of your cluster.
202

203
00:15:55,530 --> 00:15:57,150
Yeah pretty cool.
203

204
00:15:57,150 --> 00:16:04,950
And is this something where if it fails at some point it's sort of you can start where you left off
204

205
00:16:04,980 --> 00:16:06,790
if you just keep rerunning the command kind of thing.
205

206
00:16:07,590 --> 00:16:13,770
Yeah it'll try to refresh its data from the previous run and then match up to the desired state so that
206

207
00:16:13,770 --> 00:16:18,010
you are not building a whole bunch of new clusters.
207

208
00:16:18,360 --> 00:16:23,790
But if you do try to create another cluster like that was up here before and then of course it will
208

209
00:16:23,790 --> 00:16:26,410
try to create another grid.
209

210
00:16:26,560 --> 00:16:27,490
Right.
210

211
00:16:27,600 --> 00:16:28,430
Right.
211

212
00:16:28,450 --> 00:16:30,590
Yeah really cool.
212

213
00:16:30,630 --> 00:16:32,420
Anything else to show us.
213

214
00:16:32,820 --> 00:16:37,230
You know that's one of the highlights I can show you that it's it's us.
214

215
00:16:37,290 --> 00:16:44,670
It makes it so easy that you can create these clusters and get rid of them very quickly.
215

216
00:16:45,270 --> 00:16:52,960
So we just need to like try something now but don't want to pay for the cloud provider fees you know
216

217
00:16:54,150 --> 00:17:00,720
really quick to just say hey Dr. cluster removed the demo and then I'm gonna go through and start tearing
217

218
00:17:00,720 --> 00:17:01,280
it down.
218

219
00:17:02,310 --> 00:17:09,250
So it's is creating everything you need from the BPC to the subnet the security groups to the instances
219

220
00:17:09,480 --> 00:17:14,840
additional storage for your doctor volume like Marla Doctor.
220

221
00:17:15,240 --> 00:17:25,890
It's trying to use the best storage driver we can lump overly too and you know for ECP it'll set everything
221

222
00:17:25,890 --> 00:17:30,300
up that you need for DVR leave and set up like S3 object storage for you.
222

223
00:17:30,900 --> 00:17:36,120
So it works very hard to try to make it as easy as possible to get you know are our best practices in
223

224
00:17:36,120 --> 00:17:37,100
place right.
224

225
00:17:37,720 --> 00:17:38,080
Yeah.
225

226
00:17:38,250 --> 00:17:41,960
That's really cool that that was definitely in line with the goals of the DCI originally.
226

227
00:17:42,040 --> 00:17:48,630
And the reason why I like the idea of the concept a year ago he had so.
227

228
00:17:48,720 --> 00:17:49,020
All right.
228

229
00:17:49,020 --> 00:17:56,900
So I think yesterday we had been on we talked about in case you're interested in Docker desktop enterprise.
229

230
00:17:56,940 --> 00:18:04,470
This is since it is sort of the enterprise side on your on your desktop from the server enterprise stuff
230

231
00:18:04,470 --> 00:18:08,250
the docker enterprise cluster stuff that you're seeing us build here.
231

232
00:18:08,430 --> 00:18:13,050
If you have more questions about how to get get a copy of this or how to get a demo Docker desktop enterprise
232

233
00:18:13,050 --> 00:18:14,700
and stuff like that.
233

234
00:18:14,910 --> 00:18:19,320
Yesterday we talked about reaching out to your existing whoever you're already getting Docker server
234

235
00:18:19,320 --> 00:18:20,340
enterprise.
235

236
00:18:20,370 --> 00:18:23,670
I know it's like call that but that's how we have to distinguish the difference between the desktop
236

237
00:18:23,670 --> 00:18:24,940
and the server but.
237

238
00:18:25,230 --> 00:18:31,470
So if you have if you have server Docker enterprise today then you reach out to have your contact as
238

239
00:18:31,470 --> 00:18:36,780
well that's Docker directly or your partner that you're working with and you can find more out about
239

240
00:18:37,080 --> 00:18:41,610
all the new features in Docker desktop enterprise.