AWS Solution Architect Interview Questions and Answers - Part 1 — Transcript

Full transcript

1. 0:00AWS
2. 0:04architectures how would you design a
3. 0:06highly available and scalable
4. 0:08architecture for a web application that
5. 0:10is deployed on
6. 0:12ec2 we will use an elastic load balancer
7. 0:16to distribute incoming traffic across
8. 0:18multiple instances to ensure High
9. 0:21availability and fall tolerance set up
10. 0:24an autoscaling group to automatically
11. 0:26adjust the number of ec2 instances based
12. 0:29on traffic demand this ensures that the
13. 0:32application can handle wearing load
14. 0:34efficiently use multiple availability
15. 0:37zones with the autoscaling group spread
16. 0:40across them to ensure redundancy and
17. 0:43high
18. 0:44availability use Amazon elastic cache to
19. 0:47Cache frequently Access Data thereby
20. 0:49improving data read
21. 0:52performance utilize Amazon cloudfront a
22. 0:55Content delivery Network to cache and
23. 0:57deliver static and dynamic content from
24. 1:00Edge locations thus reducing content
25. 1:02delivery latency and improving user
26. 1:06experience what are good Disaster
27. 1:09Recovery strategies for a cloud
28. 1:12application our Disaster Recovery
29. 1:15strategy should be based on RTO and RPO
30. 1:18let's understand these terms first RTO
31. 1:21or recovery time objective is duration
32. 1:24within which a business process must be
33. 1:26restored RPO or recovery Point objective
34. 1:30is maximum tolerable period during which
35. 1:33data might be lost keeping these in mind
36. 1:36common Dr strategies are backup and
37. 1:39restore pilot light warm standby and
38. 1:43multi-site each varing in RTO RPO time
39. 1:46and costs
40. 1:49involved backup and restore regularly
41. 1:52backup your data and configurations
42. 1:54using aw services like Amazon S3 and
43. 1:57Amazon EBS snapshots you could use AWS
44. 2:01backup service for this ensure that
45. 2:03backups are stored in a different region
46. 2:06from your primary deployment pilot light
47. 2:09maintain a minimal version of your
48. 2:11application in a standby state in
49. 2:13another region scale up resources as
50. 2:15needed in case of a disaster use AWS
51. 2:18cloud formation for automated
52. 2:20infrastructure
53. 2:21provisioning warm standby maintain a
54. 2:24partially scaled version of your
55. 2:26application in another region with
56. 2:28essential Services run
57. 2:30and ready to scale up use AWS
58. 2:33autoscaling to manage resource scaling
59. 2:36multisite deploy your application across
60. 2:39multiple AWS regions ensuring that each
61. 2:41region can independently handle the
62. 2:43entire workload use AWS Global
63. 2:46accelerator or Route 53 for Global
64. 2:49traffic
65. 2:51management how would you design a
66. 2:53cost-effective AWS solution for
67. 2:55analyzing terabytes of historical
68. 2:57company data it should allow business
69. 2:59analyst
70. 3:00to use SQL to query and analyze data
71. 3:03relevant to their departments
72. 3:05Only Store the company's historical data
73. 3:08in Amazon S3 as a data Lake organize the
74. 3:12data into separate folders based on
75. 3:14departments to restrict access utilize
76. 3:17features like serers side encryption to
77. 3:19ensure data security use Amazon Athena
78. 3:23for analytics which would allow business
79. 3:26analysts to query and analyze data using
80. 3:30SQL for Access Control create IM roles
81. 3:33that have permissions on relevant bucket
82. 3:36subfolders using S3 prefixes allow
83. 3:39business analysts to assume the relevant
84. 3:41Department role this will give them
85. 3:43access only to the corresponding
86. 3:45Department's
87. 3:47data create an AWS architecture to
88. 3:50deliver static content to end users from
89. 3:53S3 with low latency the content to be
90. 3:56served is location dependent that is
91. 3:58different content is to be delivered to
92. 4:00users in each
93. 4:02continent organize the content in S3
94. 4:05geography wise that is in separate
95. 4:07buckets or subfolders this will serve as
96. 4:10the origins for a cloudfront
97. 4:13distribution next create a cloudfront
98. 4:15distribution with Lambda at Edge
99. 4:18function that intercepts user requests
100. 4:21and examines the HTTP header cloudfront
101. 4:23viewer country to determine the user
102. 4:25location and based on that switches the
103. 4:28origin to be used this ensures that
104. 4:31correct location specific content is
105. 4:33served via
106. 4:35cloudfront you are developing a
107. 4:37multiplayer game accessible over web and
108. 4:40mobile how would you exchange game data
109. 4:42between gamers in real
110. 4:45time you can use websockets to
111. 4:47communicate with API Gateway websockets
112. 4:50is a communication protocol that enables
113. 4:52a two-way realtime connection between a
114. 4:55web browser or mobile app client and a
115. 4:58server
116. 5:00or use a service like Amazon Kinesis
117. 5:03data streams which acts as a message
118. 5:05broker allowing clients to publish
119. 5:08updates and subscribe to receive updates
120. 5:10from other players or use appsync to
121. 5:14publish and receive messages over
122. 5:16websockets appsync is a managed
123. 5:18serverless graphql API service from AWS
124. 5:21it allows you to define a graphql schema
125. 5:24that represents your game data like
126. 5:27driver position actions
127. 5:30Etc clients can query and mutate the
128. 5:33data using graphql queries and
129. 5:37mutations a company plans to migrate its
130. 5:39on premise it setup to AWS Cloud how
131. 5:43would you plan and carry out this
132. 5:45migration AWS provides several migration
133. 5:48services to help us migrate workloads
134. 5:50from on-premise data center or other
135. 5:52Cloud providers to AWS let's see how we
136. 5:55can use
137. 5:58them the appli apption Discovery service
138. 6:00helps us collect information about
139. 6:02servers and processes in the on-premise
140. 6:05data center the migration Hub tracks and
141. 6:08manages the overall migration
142. 6:11process for a lift and shift kind of
143. 6:14migration of on premise servers to ec2
144. 6:17AWS application migration service may be
145. 6:22used for database migration DMS or
146. 6:25database migration service is
147. 6:28useful
148. 6:30and for file server copy and sync of
149. 6:32data to AWS storage like S3 or EFS AWS
150. 6:36data sync can be
151. 6:39used a logistics company wants to track
152. 6:42movement of its Fleet of trucks across
153. 6:45the country in real time on a map how
154. 6:47would you design such a
155. 6:50solution here's the overall flow every
156. 6:52truck has a Tracker device that sends
157. 6:55its identity and geographical position
158. 6:57every few seconds to our backend
159. 6:59tracking setup as a Json message the
160. 7:02tracking device could be a dedicated iot
161. 7:04device or simply the mobile phone of the
162. 7:07driver the fleet viewer application
163. 7:10receives location updates for the trucks
164. 7:12in real time from the backend Fleet
165. 7:14Tracking setup the fleet viewer
166. 7:16application will plot the location and
167. 7:19movement of the entire truck Fleet on a
168. 7:21map this is the same flow with service
169. 7:24level details location updates from the
170. 7:27trucks are sent over mqt to WS iot core
171. 7:30message broker to a specific topic and
172. 7:33iot core rule picks up the messages from
173. 7:36the topic and invokes Lambda function
174. 7:39which in turn sends the location updates
175. 7:41to Amazon location service tracker the
176. 7:44tracker saves the device location with
177. 7:46itself and generates a geol location
178. 7:49update Event Event Bridge captures the
179. 7:51event and sends it to a Lambda function
180. 7:54which sends messages to the fleet viewer
181. 7:56application clients via the API Gateway
182. 7:59or websockets and thus the fleet viewer
183. 8:01application receives and plots the truck
184. 8:04Fleet on a map note that the websocket
185. 8:06connection IDs are saved in Dynamo
186. 8:10DB a company has multiple AWS accounts
187. 8:13created over the years it wants to bring
188. 8:16all of them under one umbrella for
189. 8:18better management and Consolidated
190. 8:20billing how would you do this without
191. 8:22disturbing existing
192. 8:25operations we will use AWS organization
193. 8:28service to set up and manage multiple
194. 8:31accounts the primary account using which
195. 8:33you create an organization becomes your
196. 8:36management account under which you will
197. 8:39create organization units or Os which
198. 8:41are simply a convenient way to group
199. 8:43accounts under U you can add other
200. 8:46accounts these can be new or existing
201. 8:49accounts which you would invite to join
202. 8:51the organization setup once ready you
203. 8:54can do Consolidated billing for all
204. 8:56accounts at the organization level use
205. 8:58service control policy at organization
206. 9:01or OU level to control what services are
207. 9:03available to accounts under
208. 9:06it how would you design a system that
209. 9:09can answer natural language queries
210. 9:11using Enterprise data stored across
211. 9:13various data sources like S3 RDS jira
212. 9:19Etc the Amazon Q service is a good fit
213. 9:22for the scenario Amazon Q is an AI
214. 9:25assistant that can generate
215. 9:27comprehensive responses to your natural
216. 9:29language queries it ingests your data
217. 9:32for processing using its pre-built
218. 9:34connectors to augment its knowledge base
219. 9:37it interprets user queries refers to its
220. 9:40knowledge base and responds
221. 9:42intelligently in natural language with
222. 9:44built-in security the end users only see
223. 9:47the information that they have access
224. 9:51to a company has several rdbms databases
225. 9:54of different types one for each
226. 9:57department how can you build a data
227. 9:59wouse solution that Aggregates analyzes
228. 10:02and visualizes all the Department's data
229. 10:05together so our data sources are
230. 10:08different rdbms databases one per
231. 10:11Department we will aggregate the data in
232. 10:14red shift for this we will use DMS or
233. 10:17database migration service to replicate
234. 10:19data from The Source data sources into
235. 10:22the Target red shift data warehouse DMS
236. 10:25can do ongoing replication to keep the
237. 10:28data up to date within red shift once
238. 10:31the data is available in red shift it
239. 10:34can be queried analyzed and visualized
240. 10:36using a bi tool like quicksite