1
00:00:00,210 --> 00:00:05,970
Alone we'll come back with one more lecture on the ideas practice questions.

2
00:00:07,920 --> 00:00:13,560
What are these future that will allow you to restore your database within five minutes of a specific

3
00:00:13,560 --> 00:00:15,880
day time during that interim period.

4
00:00:16,320 --> 00:00:23,940
So let's say one of your administrators by mistake he removes part of the database or the server failed

5
00:00:24,000 --> 00:00:28,540
or you had the automated backups available.

6
00:00:30,190 --> 00:00:36,260
And Id like to restore in those because you had a problem with the main instance.

7
00:00:36,290 --> 00:00:40,820
So what you have is are these future relate to the story that is within five minutes of a specific date

8
00:00:40,820 --> 00:00:42,480
time doing that engine.

9
00:00:43,610 --> 00:00:43,910
A.

10
00:00:43,990 --> 00:00:44,610
It is.

11
00:00:44,640 --> 00:00:46,910
Are these multi easy features.

12
00:00:47,040 --> 00:00:51,010
The Multi is the feature will give me high availability not restore.

13
00:00:51,090 --> 00:00:53,360
There would be a primary and a standby.

14
00:00:53,670 --> 00:01:00,140
And if the primary fails or this will direct all the reads and writes in a standby.

15
00:01:00,210 --> 00:01:02,910
So this is not the restore part.

16
00:01:03,050 --> 00:01:07,030
Be it obvious are the manual snapshots you can cannot do.

17
00:01:07,050 --> 00:01:13,080
I mean deciding that you would like to restore to within five minutes of a specific incident or event

18
00:01:13,680 --> 00:01:14,820
or date and time.

19
00:01:14,820 --> 00:01:22,600
That means you are doing exactly your point in time the restoration or point in time recovery and maintenance

20
00:01:22,640 --> 00:01:24,680
not shots can not help you with that.

21
00:01:24,690 --> 00:01:27,520
They are not used for point in time recovery.

22
00:01:27,750 --> 00:01:28,170
Yes.

23
00:01:28,180 --> 00:01:32,340
Are these three replicas the replicas are two key.

24
00:01:32,340 --> 00:01:38,690
If I have a primary standby or even a standalone are to kill that need operations that need performance.

25
00:01:38,700 --> 00:01:41,220
So this is irrelevant in my case.

26
00:01:41,450 --> 00:01:50,130
The ABs are the point in time recovery as a feature utilizing are these automated backups and transaction

27
00:01:50,250 --> 00:01:50,740
logs.

28
00:01:50,760 --> 00:01:52,820
And this is exactly what it is.

29
00:01:53,220 --> 00:02:02,100
I can only do up to five minutes in time recovery of Restore of a database if I have the automated backups

30
00:02:02,160 --> 00:02:03,430
and the transaction logs.

31
00:02:03,450 --> 00:02:11,940
So basically the transaction logs will be applied to the automated backup snapshot that makes the most

32
00:02:11,940 --> 00:02:16,940
sense for that specific date and time or five minutes and then you can recover that.

33
00:02:17,070 --> 00:02:20,830
And if that is within the retention period you can even do better than that.

34
00:02:20,850 --> 00:02:24,230
You can do up to a few seconds.

35
00:02:24,290 --> 00:02:26,670
So that's the accuracy you can do.

36
00:02:26,840 --> 00:02:36,000
You can say that I want to recover it to Thursday to 15 and 35 seconds P-A.

37
00:02:36,080 --> 00:02:39,620
So that's how granular you can go if you you're within the detention period as well.

38
00:02:39,830 --> 00:02:44,580
So is incorrect is incorrect see is incorrect and the is the correct answer.

39
00:02:44,650 --> 00:02:45,340
A refresher.

40
00:02:45,350 --> 00:02:50,310
You can specify a point in time restore to any given second during the retention period.

41
00:02:50,390 --> 00:02:57,530
And when you initiate a point in time recovery or store the transaction logs the database transaction

42
00:02:57,530 --> 00:03:00,680
logs are applied to the most appropriate daily backup.

43
00:03:00,680 --> 00:03:07,130
Remember when you said that automated backups are done once every 24 hours on a daily basis to restore

44
00:03:07,130 --> 00:03:13,640
your database to that point in time when you restore a database instance only the default database parameters

45
00:03:13,670 --> 00:03:18,750
and security groups are associated with is restored instance so you will not get everything.

46
00:03:18,880 --> 00:03:25,480
So you need to do manual customization or just copy a configuration again once that is you need to associate

47
00:03:25,520 --> 00:03:33,420
apply the custom on the custom database parameters and security group settings right.

48
00:03:33,420 --> 00:03:38,490
Next question is are these automated backups and data be snapshots.

49
00:03:38,490 --> 00:03:44,350
For my sequel database engine are currently supported for which storage engine.

50
00:03:44,600 --> 00:03:50,560
And this is again one of the questions that you need to know the answer to and it is wrong.

51
00:03:50,570 --> 00:03:55,760
Oracle I assume it's not my ICM although my ass I mean supported but there are issues when you try to

52
00:03:55,760 --> 00:04:00,320
restore from my eyes.

53
00:04:00,500 --> 00:04:01,140
Correct.

54
00:04:01,200 --> 00:04:02,370
My sequel.

55
00:04:02,580 --> 00:04:03,080
Monk.

56
00:04:03,290 --> 00:04:05,340
So the correct answer is C.

57
00:04:05,360 --> 00:04:12,480
And again this is something that you need to know about this idea so azing correct me is incorrect C

58
00:04:12,500 --> 00:04:13,170
is correct.

59
00:04:13,190 --> 00:04:18,070
This is incorrect and your correct answer is C refresher for this question.

60
00:04:19,990 --> 00:04:28,460
Automated backups are currently supported only for the storage engine for my sequel and not for my sim

61
00:04:28,580 --> 00:04:29,180
engine.

62
00:04:31,240 --> 00:04:37,090
The point in time restore and snapshot restore features of a sequel are supported for that you need

63
00:04:37,090 --> 00:04:38,280
to be a storage engine.

64
00:04:38,280 --> 00:04:46,320
Only although my sequel supports multiple storage engines including my sem not all of them are optimized

65
00:04:46,320 --> 00:04:50,810
for crash recovery and data ability we need to restore.

66
00:04:50,940 --> 00:04:59,120
We would like to do to retain the data properly and to be able to do a recovery from a crash to the

67
00:04:59,120 --> 00:05:06,050
database in a to b is the recommended and supported storage engine for my sequel data with instance

68
00:05:06,050 --> 00:05:13,820
on Amazon Ardis next question does a charge for launch.

69
00:05:14,140 --> 00:05:20,590
But idle our DSD instances so if you have launched data with instances and you're not using them they

70
00:05:20,590 --> 00:05:23,400
are not in production they're not being tested they're not.

71
00:05:23,440 --> 00:05:26,050
There is nothing happening on them.

72
00:05:26,150 --> 00:05:28,330
Would you be charged for that.

73
00:05:28,330 --> 00:05:32,460
A This will be treated like stop two instances so no charge.

74
00:05:32,710 --> 00:05:39,770
B depends on the region and they are Diest that every instance is launched into C only if it was launched

75
00:05:39,770 --> 00:05:41,230
in the government cloud.

76
00:05:41,240 --> 00:05:47,390
Yes you will be the yes you will be charged regardless and that is the correct answer.

77
00:05:47,780 --> 00:05:51,050
It's not like you see two instances when you stop it you don't pay for it.

78
00:05:51,350 --> 00:05:56,390
As long as it is launched it is not terminated and it is then you are paying for it.

79
00:05:56,420 --> 00:05:58,460
So is wrong is wrong seems wrong.

80
00:05:58,460 --> 00:06:00,080
The is the correct answer.

81
00:06:02,230 --> 00:06:08,410
Next question which licensing types are available on E.W. as our ideas for Oracle database engine.

82
00:06:08,530 --> 00:06:17,360
So if you want to bring Oracle or you want to use Eric Oracle from E.W. us what are the licensing options

83
00:06:17,380 --> 00:06:18,730
that they have.

84
00:06:18,760 --> 00:06:19,990
Do I have to get it from me.

85
00:06:20,000 --> 00:06:20,370
Yes.

86
00:06:20,380 --> 00:06:29,180
Do I have to bring licenses a roll based license and license included license included is correct.

87
00:06:29,270 --> 00:06:33,460
Because if I'm going to take it from E.W. us but through all brazed license is wrong.

88
00:06:33,470 --> 00:06:39,530
So the quick answer is zonk be bring your own license and enterprise license.

89
00:06:39,530 --> 00:06:40,720
The first one is correct.

90
00:06:40,720 --> 00:06:42,510
Then the second half is wrong.

91
00:06:42,650 --> 00:06:43,780
And again these are facts.

92
00:06:43,790 --> 00:06:49,130
You know it or you don't not see Bring your own license and license included.

93
00:06:49,130 --> 00:06:49,760
This is correct.

94
00:06:49,760 --> 00:06:50,380
This is correct.

95
00:06:50,390 --> 00:06:56,540
And these are the two licensing models are licensing options available for Oracle on either of us public

96
00:06:56,580 --> 00:07:00,690
clouds the enterprise license unless it's included the phone is correct.

97
00:07:00,710 --> 00:07:01,890
This one is wrong.

98
00:07:02,180 --> 00:07:05,630
That's why the over all is wrong and the correct answer is C

99
00:07:11,740 --> 00:07:19,660
next question how will it w s r d s ensure a seamless fill over to the standby database instance in

100
00:07:19,660 --> 00:07:22,160
an RDF multi easy deployment deployment.

101
00:07:22,170 --> 00:07:26,740
If the primary instance fails many will feel over is initiated.

102
00:07:27,310 --> 00:07:35,200
So if you have a multi easy deployment primary and standby The question is how can it.

103
00:07:35,220 --> 00:07:43,400
Are these servers ensured a seamless spillover from the primary to the standby due to a primary that

104
00:07:43,450 --> 00:07:46,400
is instance failure or manual feel over is initiated.

105
00:07:46,400 --> 00:07:48,690
So either it is manual or automated.

106
00:07:48,770 --> 00:07:56,990
How can that be done seem as seamlessly such that the application or applications access to these databases

107
00:07:57,050 --> 00:07:59,030
will not be stopped or impacted.

108
00:07:59,030 --> 00:08:06,500
So if there was an application or applications accessing the data through the primary.

109
00:08:06,570 --> 00:08:14,690
Now how can we as our audience do they feel over such that the application access to the database although

110
00:08:14,690 --> 00:08:21,470
it's going to be from the standby the application will not see any difference that has happened during

111
00:08:21,470 --> 00:08:27,740
the fall over from the primary to stand by and the application access database will continue and you

112
00:08:27,740 --> 00:08:36,260
need to choose to answers a day that his audience will have an internal IP which will read direct requests

113
00:08:36,260 --> 00:08:37,010
to the new data.

114
00:08:37,030 --> 00:08:42,470
So this is talking as if the R S is an L B that will have an IP address.

115
00:08:42,470 --> 00:08:46,440
The application will point to that IP address and then it will either go to primary or secondary.

116
00:08:46,440 --> 00:08:48,550
This is not how it works.

117
00:08:48,560 --> 00:08:49,650
So this is wrong.

118
00:08:49,780 --> 00:08:55,380
Be intelligence already relies on DNS to switch over to the standby seamlessly.

119
00:08:55,380 --> 00:08:59,610
This is correct although not fully explained in this option but that's correct.

120
00:08:59,930 --> 00:09:09,490
And basically what it does is it points to the endpoint of the Multi is the audience intimidation and

121
00:09:09,490 --> 00:09:12,690
the end point is the same for the primary and stand.

122
00:09:12,820 --> 00:09:21,190
However the end point when the primary is up the audience will populate the DNS entry for the end point

123
00:09:21,190 --> 00:09:28,960
to point to the IP address of the primary if the primary fails over to the secondary then it will that

124
00:09:29,050 --> 00:09:35,650
the community can name or the seen him record in the DNS will point Ardelia's will change it to point

125
00:09:35,650 --> 00:09:43,760
to the secondary instance or the Stand by instance and not the primary see it up here.

126
00:09:43,780 --> 00:09:49,310
Our audience will have both primary and standby and independently switch over must be done manually.

127
00:09:49,320 --> 00:09:54,200
Of course this is not what we learned for what we know about our audience.

128
00:09:54,260 --> 00:09:58,880
We know that they are in primary standby and there is a continuous synchronization and the primary will

129
00:09:58,880 --> 00:10:04,350
be the source of reads and writes the standby it is not for read or write until a field over happens.

130
00:10:04,700 --> 00:10:08,560
So they're not running independently at all and that's why the answer is wrong.

131
00:10:09,020 --> 00:10:14,270
The scene him or the canonical name record is updated from primary standby.

132
00:10:14,270 --> 00:10:21,740
Basically they're seen in record in the DNS will point initially that the endpoint will map to the IP

133
00:10:21,740 --> 00:10:28,950
address of the primary when the switch over happens then it will point to the IP of the.

134
00:10:29,000 --> 00:10:31,080
Exactly standby.

135
00:10:31,160 --> 00:10:39,110
So be is correct so the correct answers are D and D quickly official for the scope doing feel over the

136
00:10:39,110 --> 00:10:44,810
scene seam of the audience that instance is updated to map to the stand by IP address instead of the

137
00:10:44,870 --> 00:10:47,190
primary IP address.

138
00:10:47,210 --> 00:10:52,730
This is why it is always recommended if you want the application not to be impacted or affected that

139
00:10:52,730 --> 00:10:59,330
the applications must point to the end point of the database and not to the IP address of the primary

140
00:10:59,360 --> 00:11:00,880
or standby.

141
00:11:01,700 --> 00:11:05,210
But if you point if the application points to the IP address will it work.

142
00:11:05,210 --> 00:11:10,460
Yes but if a film ever happens you need to go into the application and change the IP address to be done

143
00:11:10,470 --> 00:11:16,540
by IP address the name does not change because the audience is in point does not change.

144
00:11:16,550 --> 00:11:19,780
However the mapping to the IP address changes.

145
00:11:19,880 --> 00:11:27,940
Right so last question our audience your client has a global presence with offices all over the world.

146
00:11:27,980 --> 00:11:34,110
So your client has offices all over the world all connected built in different regions.

147
00:11:34,370 --> 00:11:41,810
ASL is tracking three tier application is used in every region leveraging multi Izzi my secret RDL database

148
00:11:41,840 --> 00:11:49,100
instances the client has a requirement to build an analytics based reporting application in the HQ region

149
00:11:49,610 --> 00:11:55,520
that will need to do intensive queries to the regional databases to collect information and generate

150
00:11:55,520 --> 00:11:56,580
reports.

151
00:11:57,110 --> 00:11:58,720
So what are we saying.

152
00:11:58,730 --> 00:12:00,350
This is a long story now.

153
00:12:00,620 --> 00:12:07,730
So we have multiple regions and this company is a global company it has a headquarter that is deployed

154
00:12:07,760 --> 00:12:15,590
in the region but also has different regions across the globe where they have a sales tracking or a

155
00:12:15,590 --> 00:12:18,510
sales reporting three tiered application.

156
00:12:18,690 --> 00:12:26,550
Okay could be that data entry is happening all the time on the web layer of the application and then

157
00:12:26,550 --> 00:12:27,500
stored in the database.

158
00:12:27,510 --> 00:12:32,130
Anyways I have here my sequel multi that is in every region.

159
00:12:32,140 --> 00:12:39,210
So I have primary standby in every region and every region is responsible for collecting its own information

160
00:12:39,260 --> 00:12:39,920
fine.

161
00:12:40,320 --> 00:12:41,480
So what do you need to do now.

162
00:12:41,700 --> 00:12:49,320
I need to run analytics reporting application and the application needs to intensive to do intensive

163
00:12:49,320 --> 00:12:57,870
reads queries to the different regional implementations of the database collect information out of them

164
00:12:57,960 --> 00:13:06,680
and then generate reports to the top or the higher management in the headquarter the gushingly Humvee's

165
00:13:08,350 --> 00:13:15,790
How can you architect the database access database access and the analytics to to query a database such

166
00:13:15,790 --> 00:13:18,930
that the original database performance is not impacted.

167
00:13:18,940 --> 00:13:24,030
Now I have other applications may be happening here and querying the database locally.

168
00:13:24,310 --> 00:13:27,960
Now I don't want these to be impacted by the intensive queries.

169
00:13:27,960 --> 00:13:31,990
I will do by the analytical analytics for application.

170
00:13:32,080 --> 00:13:35,660
The solution is cost effective and the reports can be done quickly.

171
00:13:35,680 --> 00:13:43,090
So these are the requirements in place the sent by instance for each region in the HQ region.

172
00:13:43,120 --> 00:13:49,630
I cannot do that because we know that the primary and standby for a multi is the database or are the

173
00:13:49,630 --> 00:13:53,210
US must be in the same region different availability zones.

174
00:13:53,380 --> 00:13:59,340
We create our snapshots and save them in an S3 bucket in the HQ region.

175
00:13:59,360 --> 00:14:07,290
So here they are talking about manually creating snapshots of the database and saving them in an as

176
00:14:07,300 --> 00:14:07,700
three.

177
00:14:07,720 --> 00:14:11,610
But the question is if you have the snapshots What can you do with them.

178
00:14:11,670 --> 00:14:12,570
You're not going.

179
00:14:12,760 --> 00:14:17,140
Each time you want to restore from a snapshot you are creating and you that is instance and if you do

180
00:14:17,140 --> 00:14:25,570
that hourly It wouldn't make sense that I'm going to create 24 instances of the 24 hour snapshots that

181
00:14:25,570 --> 00:14:28,660
I'm going to take hourly every day.

182
00:14:28,690 --> 00:14:32,380
So this is nonsense and that's why I will rule this one out as well.

183
00:14:32,770 --> 00:14:33,780
See.

184
00:14:33,910 --> 00:14:40,460
Use Google dial connect lengths to connect all regional sequel deployments to the HQ region and run

185
00:14:40,660 --> 00:14:43,040
the database queries over DX links.

186
00:14:43,090 --> 00:14:53,260
So is suggesting that you create direct connect links from the HQ to every single region globally which

187
00:14:53,260 --> 00:14:55,030
is heaps and heaps of money.

188
00:14:55,030 --> 00:14:57,330
It is very very very expensive.

189
00:14:57,580 --> 00:15:00,900
And then use the direct connect links to query that.

190
00:15:01,030 --> 00:15:02,690
So this one has two problems.

191
00:15:02,700 --> 00:15:07,900
The first problem is the region and that there is performance will be impacted because I'm quitting

192
00:15:07,900 --> 00:15:15,970
doing that intensive queries on the database itself and also is not cost effective.

193
00:15:16,030 --> 00:15:19,340
So I am losing two of the requirements that I am going to rule this one out.

194
00:15:19,360 --> 00:15:22,000
Let's look at the for each region.

195
00:15:22,000 --> 00:15:25,260
Create across the region the replica in the HQ region.

196
00:15:25,270 --> 00:15:33,490
So from each region since you have primary in standby create a replica from each region into exactly

197
00:15:34,120 --> 00:15:36,160
into the headquarter region.

198
00:15:36,160 --> 00:15:45,490
So that becomes makes it very close to the HQ analytics application so that can quickly query the databases

199
00:15:45,510 --> 00:15:46,320
locally.

200
00:15:46,360 --> 00:15:54,480
And these are really replicas so I'm not impacting the original or the original performance of the databases.

201
00:15:54,730 --> 00:16:00,360
It is cost effective because I'm creating only there because now and it can be done quickly.

202
00:16:00,390 --> 00:16:05,680
So compared to the other options this one seems to be the one that makes perfect sense.

203
00:16:05,680 --> 00:16:11,220
So Jerry's analysis is wrong be wrong seems wrong and D is the correct answer.

204
00:16:11,220 --> 00:16:11,710
All right.

205
00:16:11,710 --> 00:16:19,270
So remember that read replicas are very handy when you are talking about cross the gender application

206
00:16:19,270 --> 00:16:26,060
when you are talking about Diyar and you want your analytics application to do quitting intensive quitting

207
00:16:26,090 --> 00:16:31,840
grid operations on the database and you don't want the death of Jesus to be impacted version of the

208
00:16:31,840 --> 00:16:33,180
beast to be impacted.

209
00:16:33,190 --> 00:16:40,870
All right so that brings us to the end of the audience examplar difficulty level practice questions

210
00:16:40,990 --> 00:16:45,320
and I hope now RLDS is much easier for you.

211
00:16:45,340 --> 00:16:51,670
This practice had grouped a lot of ideas that you might see in the exam or you might see very close

212
00:16:51,670 --> 00:16:54,780
ideas in the exam and I think we have covered the full nine yards.

213
00:16:54,850 --> 00:17:00,160
Whichever you answered try to associate that you are doing that requires our D.S. competency I think

214
00:17:00,160 --> 00:17:05,620
you have reached a very high competency level by doing this practice which is very important.

215
00:17:05,830 --> 00:17:11,740
OK so until we meet in the next section please be safe and I will see then.