WEBVTT

1
00:00.000 --> 00:07.660
So now that we have a good understanding of why we need good performance and

2
00:07.660 --> 00:08.680
redundancy

3
00:08.680 --> 00:13.000
with connectivity to our cloud, and we've talked a little bit about some of the

4
00:13.000 --> 00:13.680
different

5
00:13.680 --> 00:18.920
characteristics that we need with that redundancy and performance, let's jump

6
00:18.920 --> 00:19.760
in and take a

7
00:19.760 --> 00:23.440
look at some of these things in a little more detail.

8
00:23.440 --> 00:26.880
We're going to be focusing here with the on-prem equipment.

9
00:26.880 --> 00:31.800
However, keep in mind that if you're deploying your own routers into the cloud
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

9
00:26.880 --> 00:31.800
However, keep in mind that if you're deploying your own routers into the cloud

10
00:31.800 --> 00:32.680
environment,

11
00:32.680 --> 00:37.480
a lot of these same things may apply to that equipment as well.

12
00:37.480 --> 00:42.620
And we may be using the same technologies on both sides or in conjunction with

13
00:42.620 --> 00:43.840
each other.

14
00:43.840 --> 00:47.890
So let's go ahead and jump in and take a closer look at some of these different

15
00:47.890 --> 00:48.720
technologies

16
00:48.720 --> 00:53.840
and aspects of performance and resiliency that we need to provide with our

17
00:53.840 --> 00:55.400
cloud connectivity.

18
00:56.400 --> 01:02.610
So as we jump in and we start exploring our redundancy requirements, what we
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

18
00:56.400 --> 01:02.610
So as we jump in and we start exploring our redundancy requirements, what we

19
01:02.610 --> 01:03.160
want to do

20
01:03.160 --> 01:07.560
is sort of focus in on what are the critical design points.

21
01:07.560 --> 01:11.640
What are the things that we have to make sure that we're including in order to

22
01:11.640 --> 01:12.360
provide the

23
01:12.360 --> 01:16.160
redundancy that we need for mission critical applications?

24
01:16.160 --> 01:21.280
The first, of course, is going to be redundant internet connections.

25
01:21.280 --> 01:25.160
Now, what that actually means could, of course, vary.

26
01:25.160 --> 01:27.680
It means we could have a primary internet.

27
01:27.680 --> 01:31.480
Maybe we have something like a 5G backup connection.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

27
01:27.680 --> 01:31.480
Maybe we have something like a 5G backup connection.

28
01:31.480 --> 01:35.190
Maybe it means that we have a direct internet connection and then we have a

29
01:35.190 --> 01:36.320
private connection

30
01:36.320 --> 01:40.360
through a co-location that can also go out to the internet.

31
01:40.360 --> 01:46.050
But we can't have redundancy going to the cloud, particularly, of course,

32
01:46.050 --> 01:47.160
public cloud

33
01:47.160 --> 01:51.990
providers if we don't have redundant internet connections or at least an

34
01:51.990 --> 01:53.480
internet connection

35
01:53.480 --> 01:56.160
and a private connection as well.

36
01:56.160 --> 02:02.360
So some combination of redundant paths to get to the cloud provider.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

36
01:56.160 --> 02:02.360
So some combination of redundant paths to get to the cloud provider.

37
02:02.360 --> 02:08.280
Often an important aspect of that is to also have redundant service providers.

38
02:08.280 --> 02:12.360
You don't want to get both of these connections through the same service

39
02:12.360 --> 02:13.240
provider.

40
02:13.240 --> 02:17.240
If that service provider were to fail, well, then both of your connections is

41
02:17.240 --> 02:17.760
going to

42
02:17.760 --> 02:18.800
fail.

43
02:18.800 --> 02:23.760
When we say redundant connections, oftentimes, we're talking about the last

44
02:23.760 --> 02:25.160
mile, the connection

45
02:25.160 --> 02:28.480
between the service provider and your company.

46
02:28.480 --> 02:33.810
Well, again, that's fine, but what if it's the service provider itself that
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

46
02:28.480 --> 02:33.810
Well, again, that's fine, but what if it's the service provider itself that

47
02:33.810 --> 02:34.600
fails?

48
02:34.600 --> 02:38.830
That's why we want to make sure we get connectivity from two different service

49
02:38.830 --> 02:39.720
providers.

50
02:39.720 --> 02:41.920
And just a little side note on that.

51
02:41.920 --> 02:47.600
We need to make sure that we actually have parent redundancy.

52
02:47.600 --> 02:51.230
In case you're not familiar with exactly what that means, let me just give you

53
02:51.230 --> 02:52.600
an example.

54
02:52.600 --> 02:57.990
I had a customer one time who had a primary internet connection through a major

55
02:57.990 --> 02:58.640
service

56
02:58.640 --> 02:59.640
provider.

57
02:59.640 --> 03:02.000
I'm not going to name names here.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

57
02:59.640 --> 03:02.000
I'm not going to name names here.

58
03:02.000 --> 03:05.940
Well one time that connection failed and they decided they wanted to get a

59
03:05.940 --> 03:07.520
redundant connection

60
03:07.520 --> 03:11.780
and they decided to follow this goal of going through two different service

61
03:11.780 --> 03:12.680
providers.

62
03:12.680 --> 03:19.280
So they got a secondary connection through sort of a lower tier local ISP.

63
03:19.280 --> 03:25.790
Well what they didn't know is that that ISP also went through their primary ISP

64
03:25.790 --> 03:26.080
to get

65
03:26.080 --> 03:27.760
out to the internet.

66
03:27.760 --> 03:28.760
So guess what?

67
03:28.760 --> 03:34.230
About a year and a half later, their primary service provider had an outage
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

67
03:28.760 --> 03:34.230
About a year and a half later, their primary service provider had an outage

68
03:34.230 --> 03:34.880
again and big

69
03:34.880 --> 03:40.300
surprise their backup didn't work either because again the backup service

70
03:40.300 --> 03:41.040
provider

71
03:41.040 --> 03:46.160
was using their primary service provider just like they were.

72
03:46.160 --> 03:50.040
So that's what we mean by having parent redundancy.

73
03:50.040 --> 03:55.400
Make sure that your second service provider is not also a customer of your

74
03:55.400 --> 03:56.560
first service

75
03:56.560 --> 04:01.840
provider because that happens a lot with like tier one, tier two and tier three
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

75
03:56.560 --> 04:01.840
provider because that happens a lot with like tier one, tier two and tier three

76
04:01.840 --> 04:04.440
service providers.

77
04:04.440 --> 04:07.690
The way you would know that by the way is you would simply have to ask the

78
04:07.690 --> 04:08.280
secondary

79
04:08.280 --> 04:10.000
service provider.

80
04:10.000 --> 04:14.700
And of course as we sort of mentioned earlier having redundant connections be

81
04:14.700 --> 04:15.480
them to redundant

82
04:15.480 --> 04:20.430
ISPs or not doesn't do us a whole lot of good if we don't have redundant

83
04:20.430 --> 04:21.640
equipment.

84
04:21.640 --> 04:26.470
So redundant edge routers, redundant switches within the company to ensure that

85
04:26.470 --> 04:27.280
we're not

86
04:27.280 --> 04:32.480
relying on any single device that's going to create a single point of failure.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

86
04:27.280 --> 04:32.480
relying on any single device that's going to create a single point of failure.

87
04:32.480 --> 04:36.760
Those redundant devices of course could also have redundant components.

88
04:36.760 --> 04:41.360
We can certainly have both things like redundant power supplies.

89
04:41.360 --> 04:45.320
This is probably the biggest one to be honest with you.

90
04:45.320 --> 04:51.030
Most equipment out there that's enterprise level equipment has a solution for

91
04:51.030 --> 04:51.880
dual power

92
04:51.880 --> 04:53.280
supplies.

93
04:53.280 --> 04:58.590
Even if on some of the smaller equipment that might be an external power supply

94
04:58.590 --> 04:59.240
and then

95
04:59.240 --> 05:03.560
there's the internal which is the main power supply but there's usually some
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

95
04:59.240 --> 05:03.560
there's the internal which is the main power supply but there's usually some

96
05:03.560 --> 05:04.480
way to provide

97
05:04.480 --> 05:06.760
redundant power.

98
05:06.760 --> 05:13.560
Also redundant supervisor engines for switches or route processors for routers.

99
05:13.560 --> 05:17.720
Now certainly not all of the equipment supports this option.

100
05:17.720 --> 05:22.280
This is generally going to be again enterprise class equipment but if the

101
05:22.280 --> 05:23.440
equipment supports

102
05:23.440 --> 05:26.560
it these are always a great choice.

103
05:26.560 --> 05:31.420
In the same thing sort of with fan modules, redundant fan modules honestly used
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

103
05:26.560 --> 05:31.420
In the same thing sort of with fan modules, redundant fan modules honestly used

104
05:31.420 --> 05:31.920
to just

105
05:31.920 --> 05:37.020
be for the higher end equipment but redundancy on the fan modules is actually

106
05:37.020 --> 05:37.840
coming down

107
05:37.840 --> 05:42.750
to the smaller lower end equipment as well where oftentimes what they'll do is

108
05:42.750 --> 05:43.360
they'll

109
05:43.360 --> 05:48.600
put the fans in parallel so one fan is actually blowing through a second fan.

110
05:48.600 --> 05:53.730
That way if either one of the two fans fail that particular we'll call it air

111
05:53.730 --> 05:54.360
path is

112
05:54.360 --> 05:56.400
still pushing air through.

113
05:56.400 --> 06:00.660
So the redundancy is actually back to back fans in some of the smaller
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

113
05:56.400 --> 06:00.660
So the redundancy is actually back to back fans in some of the smaller

114
06:00.660 --> 06:01.720
equipment that

115
06:01.720 --> 06:04.960
doesn't have a great big fan tray on the side.

116
06:04.960 --> 06:08.610
Now just focusing in on our redundant power supplies for a moment they're not

117
06:08.610 --> 06:09.040
going to

118
06:09.040 --> 06:15.440
do any good unless we also have redundant power going to the device itself.

119
06:15.440 --> 06:19.130
So we can plug in both power supplies but we need to make sure we have

120
06:19.130 --> 06:20.240
redundant power

121
06:20.240 --> 06:22.600
to our facility as well.

122
06:22.600 --> 06:25.480
Some of these are going to be far easier to achieve than others but let's go

123
06:25.480 --> 06:25.960
through

124
06:25.960 --> 06:27.760
our list.

125
06:27.760 --> 06:32.760
Probably the hardest one to do is to have redundant power grids.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

125
06:27.760 --> 06:32.760
Probably the hardest one to do is to have redundant power grids.

126
06:32.760 --> 06:36.770
The reason this one is going to be the hardest is this one of course is up to

127
06:36.770 --> 06:39.240
the power company.

128
06:39.240 --> 06:43.630
Whether the power company can actually connect you to redundant power grids and

129
06:43.630 --> 06:44.720
a lot of that

130
06:44.720 --> 06:48.760
is going to depend on location.

131
06:48.760 --> 06:53.350
So again depending on the power company and your location this is going to be a

132
06:53.350 --> 06:53.800
great

133
06:53.800 --> 06:58.040
big question mark of whether you can actually do this one.

134
06:58.040 --> 07:02.540
One of the enterprises that I've done a lot of work for actually chose where to
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

134
06:58.040 --> 07:02.540
One of the enterprises that I've done a lot of work for actually chose where to

135
07:02.540 --> 07:03.000
build

136
07:03.000 --> 07:08.190
their data center based on whether they could connect to redundant power grids

137
07:08.190 --> 07:08.920
or not.

138
07:08.920 --> 07:13.400
So they actually chose the location based on availability.

139
07:13.400 --> 07:17.340
So of all the things we're going to mention here this is probably the trickiest

140
07:17.340 --> 07:17.520
because

141
07:17.520 --> 07:21.640
again it's not usually going to be under your control.

142
07:21.640 --> 07:26.790
Now keep in mind things like redundant power grids are also going to depend a

143
07:26.790 --> 07:27.520
lot on the

144
07:27.520 --> 07:29.320
size of your company.

145
07:29.320 --> 07:34.570
For example let's say Amazon was going to build a new data center in your town
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

145
07:29.320 --> 07:34.570
For example let's say Amazon was going to build a new data center in your town

146
07:34.570 --> 07:34.800
and they

147
07:34.800 --> 07:39.590
went to the local power company and said look we want connections to multiple

148
07:39.590 --> 07:40.400
grids and

149
07:40.400 --> 07:43.520
we're building our data center here.

150
07:43.520 --> 07:48.690
My guess would be the power company would probably say we can run some lines

151
07:48.690 --> 07:49.400
for you.

152
07:49.400 --> 07:53.680
We can definitely get you connected into multiple power grids.

153
07:53.680 --> 07:57.470
Where of course on the other hand if a smaller company came in and said that

154
07:57.470 --> 07:57.920
their answer

155
07:57.920 --> 08:02.200
would be now can't connect you to multiple power grids at that location.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

155
07:57.920 --> 08:02.200
would be now can't connect you to multiple power grids at that location.

156
08:02.200 --> 08:07.930
So as fair or unfair as it may be in this particular case the size of the

157
08:07.930 --> 08:09.560
company probably

158
08:09.560 --> 08:13.780
is going to have a bit to do with whether you can pull off this connection if

159
08:13.780 --> 08:14.360
it's not

160
08:14.360 --> 08:19.080
just readily available where you choose to put your data center.

161
08:19.080 --> 08:21.880
Now let's take a look at some of our easier options.

162
08:21.880 --> 08:24.760
We can certainly have redundant UPS's.

163
08:24.760 --> 08:26.600
This one's relatively easy.

164
08:26.600 --> 08:31.900
We simply take the line power run it into two different UPS's and then each UPS
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

164
08:26.600 --> 08:31.900
We simply take the line power run it into two different UPS's and then each UPS

165
08:31.900 --> 08:32.320
goes

166
08:32.320 --> 08:36.000
into our redundant power supplies in our equipment.

167
08:36.000 --> 08:38.440
So this one's relatively easy.

168
08:38.440 --> 08:44.430
Some that's often overlooked is automatic transfer switches or I've always sort

169
08:44.430 --> 08:45.120
of called them

170
08:45.120 --> 08:47.240
high voltage switches.

171
08:47.240 --> 08:52.770
But if you're not familiar with this automatic transfer switches or ATS's are

172
08:52.770 --> 08:53.800
what connect

173
08:53.800 --> 08:59.440
your equipment so your data center power to the line power.

174
08:59.440 --> 09:03.440
So this is what's connecting you to these power grids that we were just talking
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

174
08:59.440 --> 09:03.440
So this is what's connecting you to these power grids that we were just talking

175
09:03.440 --> 09:03.440
about

176
09:03.440 --> 09:09.190
and these switches sit between the power grid and our next choice which is to

177
09:09.190 --> 09:11.360
have generators.

178
09:11.360 --> 09:17.020
So these ATS devices sit between the power grid and the generator and when your

179
09:17.020 --> 09:17.440
line

180
09:17.440 --> 09:19.240
power fails.

181
09:19.240 --> 09:25.990
So if this power were to go off the ATS kicks in and switches you over to the

182
09:25.990 --> 09:27.480
generators.

183
09:27.480 --> 09:29.080
That's what this does.

184
09:29.080 --> 09:34.880
Well the problem can be if you only have one set of ATS's that can be your
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

184
09:29.080 --> 09:34.880
Well the problem can be if you only have one set of ATS's that can be your

185
09:34.880 --> 09:36.040
single point

186
09:36.040 --> 09:42.400
of failure and you don't switch over to your generators if the ATS fails.

187
09:42.400 --> 09:47.010
Now also keep in mind since it's on our list here we can also have redundant

188
09:47.010 --> 09:47.880
generators

189
09:47.880 --> 09:51.160
and oftentimes if you're going to have redundant generators you're probably

190
09:51.160 --> 09:52.960
going to want redundant

191
09:52.960 --> 09:54.600
fuel supplies.

192
09:54.600 --> 09:59.490
And again just to throw out an example here from my experience one of the large

193
09:59.490 --> 10:00.360
enterprises
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

193
09:59.490 --> 10:00.360
enterprises

194
10:00.360 --> 10:06.200
that I've worked with had everything on this list except one thing.

195
10:06.200 --> 10:10.960
They actually had redundant power grids to their data center.

196
10:10.960 --> 10:15.760
They had redundant UPS's at the end of every rack.

197
10:15.760 --> 10:20.760
They had redundant generators with redundant fuel supplies.

198
10:20.760 --> 10:26.080
So we had redundant diesel tanks connected to our generators but guess what?

199
10:26.080 --> 10:30.240
They only had one set of high voltage switching gear.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

199
10:26.080 --> 10:30.240
They only had one set of high voltage switching gear.

200
10:30.240 --> 10:34.920
Would anybody out there like to take a guess as to what failed that took down

201
10:34.920 --> 10:35.760
our data center

202
10:35.760 --> 10:38.040
for about eight hours?

203
10:38.040 --> 10:40.080
I don't think I need to answer that.

204
10:40.080 --> 10:41.960
I think you get the idea.

205
10:41.960 --> 10:47.710
I could also ask you what do you think was purchased the next week after the

206
10:47.710 --> 10:48.200
CEO of the

207
10:48.200 --> 10:51.480
company found out what happened.

208
10:51.480 --> 10:57.840
We very quickly installed a second set of high voltage switching gear.

209
10:57.840 --> 11:01.920
Now our next topic here is definitely something we want to be concerned with.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

209
10:57.840 --> 11:01.920
Now our next topic here is definitely something we want to be concerned with.

210
11:01.920 --> 11:06.430
Having redundant cooling doesn't do us any good if everything keeps running but

211
11:06.430 --> 11:07.200
it overheats.

212
11:07.200 --> 11:12.110
So we definitely want redundant cooling but I will say that this is something

213
11:12.110 --> 11:13.440
that's actually

214
11:13.440 --> 11:20.150
usually already taken care of or already covered just by the design of the HVAC

215
11:20.150 --> 11:21.280
systems.

216
11:21.280 --> 11:24.160
There's usually multiple systems.

217
11:24.160 --> 11:29.230
The customer I was just talking about with this data center up here, they had

218
11:29.230 --> 11:31.040
14 air exchangers
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

218
11:29.230 --> 11:31.040
14 air exchangers

219
11:31.040 --> 11:32.760
in their data center.

220
11:32.760 --> 11:38.410
So two rows of seven up the outside walls of the data center and the way it was

221
11:38.410 --> 11:39.280
designed

222
11:39.280 --> 11:44.360
is the data center could actually be kept cool enough with only 12.

223
11:44.360 --> 11:47.960
So they only needed 12 systems but they had 14.

224
11:47.960 --> 11:54.040
So again we could have lost two complete HVAC systems and they would have been

225
11:54.040 --> 11:55.040
covered.

226
11:55.040 --> 11:59.180
Those HVAC systems were completely independent from each other except for the

227
11:59.180 --> 11:59.800
fact that they

228
11:59.800 --> 12:02.520
were controlled by the central unit.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

228
11:59.800 --> 12:02.520
were controlled by the central unit.

229
12:02.520 --> 12:06.510
In other words you could still set the temperature and humidity and so on on a

230
12:06.510 --> 12:07.440
central control

231
12:07.440 --> 12:08.520
unit.

232
12:08.520 --> 12:11.940
Now something else we need to be concerned with is all of this redundancy we're

233
12:11.940 --> 12:13.560
talking about,

234
12:13.560 --> 12:15.880
none of this will do us any good.

235
12:15.880 --> 12:21.010
So all of these redundant devices and connections, they do us no good at all

236
12:21.010 --> 12:22.160
unless we actually

237
12:22.160 --> 12:28.140
detect a failure and can therefore implement a switch over to the redundant

238
12:28.140 --> 12:29.160
devices.

239
12:29.160 --> 12:33.720
Again, we have to have this in place or nothing's going to do any good.
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

239
12:29.160 --> 12:33.720
Again, we have to have this in place or nothing's going to do any good.

240
12:33.720 --> 12:35.320
Well how do we do that?

241
12:35.320 --> 12:39.920
There's actually a ton of ways, one of them is dynamic routing.

242
12:39.920 --> 12:46.280
So things like OSPF, EIGRP, I mean I wouldn't recommend RIP necessarily but

243
12:46.280 --> 12:47.400
really any sort

244
12:47.400 --> 12:53.560
of dynamic routing and technically you could do it with static routing too.

245
12:53.560 --> 12:57.520
You can actually implement something called a floating static route to act as a

246
12:57.520 --> 12:57.920
backup

247
12:57.920 --> 13:02.530
to your primary path and you can do it with static but usually we're going to
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

247
12:57.920 --> 13:02.530
to your primary path and you can do it with static but usually we're going to

248
13:02.530 --> 13:03.100
be using

249
13:03.100 --> 13:04.740
dynamic routing.

250
13:04.740 --> 13:09.800
You can do this with IPSLA monitors to actually monitor our links.

251
13:09.800 --> 13:15.600
We can have IPSLA monitors, trigger tracked objects to affect a failover.

252
13:15.600 --> 13:19.240
We can do this with bidirectional forwarding detection.

253
13:19.240 --> 13:24.360
This can detect a failure in as little as 150 milliseconds.

254
13:24.360 --> 13:29.410
So this is going to be a high speed way to detect failure even faster than our

255
13:29.410 --> 13:30.080
dynamic
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

255
13:29.410 --> 13:30.080
dynamic

256
13:30.080 --> 13:31.320
routing.

257
13:31.320 --> 13:35.420
The dynamic routing and the bidirectional forwarding detection are often going

258
13:35.420 --> 13:35.860
to work

259
13:35.860 --> 13:37.980
very closely together.

260
13:37.980 --> 13:41.340
Dynamic routing is going to monitor the BFD.

261
13:41.340 --> 13:48.530
Also first hop redundancy protocols so again things like VRRP, HSRP and GLBP if

262
13:48.530 --> 13:49.420
you want

263
13:49.420 --> 13:53.990
to do gateway load balancing protocol and get a little bit of load sharing out

264
13:53.990 --> 13:54.580
of it.

265
13:54.580 --> 13:58.900
Also don't forget things like chassis clustering.

266
13:58.900 --> 14:05.050
Things like our Cisco firewalls support clustering, things like stack wise,
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

266
13:58.900 --> 14:05.050
Things like our Cisco firewalls support clustering, things like stack wise,

267
14:05.050 --> 14:05.980
where we have redundant

268
14:05.980 --> 14:11.110
switches acting as a single unit and sort of the predecessor to stack wise

269
14:11.110 --> 14:12.100
virtual is

270
14:12.100 --> 14:17.370
VSS or the virtual switching system which was also a way of basically clust

271
14:17.370 --> 14:18.560
ering together

272
14:18.560 --> 14:22.300
multiple devices to act in an redundant pair.

273
14:22.300 --> 14:27.220
Now we don't really have a good way to explore the chassis clustering.

274
14:27.220 --> 14:32.410
However these options up here at the top, the dynamic routing, the IPSLA
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

274
14:27.220 --> 14:32.410
However these options up here at the top, the dynamic routing, the IPSLA

275
14:32.410 --> 14:33.340
monitors, the

276
14:33.340 --> 14:39.770
tracked objects, BFD and first hop redundancy protocols will be exploring all

277
14:39.770 --> 14:40.700
of these in

278
14:40.700 --> 14:45.260
much greater detail in some of our upcoming skills.

279
14:45.260 --> 14:49.540
So you'll see a lot more on these topics a little bit later.

280
14:49.540 --> 14:55.480
So in this video we took a closer look at some of our failover or redundancy

281
14:55.480 --> 14:56.460
options that

282
14:56.460 --> 15:02.640
we have mostly that we can use for our on-prem equipment or our local data
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

282
14:56.460 --> 15:02.640
we have mostly that we can use for our on-prem equipment or our local data

283
15:02.640 --> 15:03.780
center.

284
15:03.780 --> 15:08.480
Now that said, can some of these things be used in conjunction with the cloud

285
15:08.480 --> 15:09.300
services?

286
15:09.300 --> 15:10.860
Of course they can.

287
15:10.860 --> 15:14.400
And if you're running your own routers in the cloud, we could very easily do

288
15:14.400 --> 15:14.860
some of

289
15:14.860 --> 15:16.100
these features.

290
15:16.100 --> 15:20.140
However, this was again mostly focusing on-prem.

291
15:20.140 --> 15:25.320
Next, let's take a look at some of the things offered by the various cloud

292
15:25.320 --> 15:26.420
providers.

293
15:26.420 --> 15:29.500
Let's start with AWS in the next video.

294
15:29.500 --> 15:37.520
[BLANK_AUDIO]
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:900000

294
15:29.500 --> 15:37.520
[BLANK_AUDIO]