AIAI Agent Dev Labaiagentdev-lab.com
Sign in to save progressSign in
Menu

CLF-C02 · LESSON 3

Cloud Technology and Services34% · 8 topics

Learn methods of deploying and operating on AWS, the global infrastructure, and compute, database, networking, storage, AI/ML, analytics and other services.

3.1Understand methods of deploying and operating in the AWS Cloud

3.1 Define methods of deploying and operating in the AWS Cloud.

Concept

You can access AWS through the Management Console (UI), the AWS CLI (commands) and SDKs (programs). AWS CloudFormation is used for IaC, defining infrastructure as code. Connectivity options are VPN over the internet (AWS Site-to-Site VPN) and dedicated lines with AWS Direct Connect. Deployment models are cloud, hybrid and on-premises (AWS Outposts and so on).

How it works / steps

  1. Use the console for one-off operations and automate repeated work with the CLI or SDKs
  2. Make environment builds reproducible with CloudFormation templates
  3. For on-premises connectivity, choose quick-to-use VPN or stable dedicated Direct Connect
  4. Consider Outposts if you need AWS infrastructure on-premises

Workplace example

Staging and production environments are built from the same CloudFormation template, and the headquarters data center is connected with Direct Connect.

Common confusions and exam tips

Mini quiz

Q1. The same procedure is repeated every month in each department’s AWS account. What should be done next to reduce errors?

  1. Record results with Config and find mistakes later
  2. Assign one person permanently so practice reduces mistakes
  3. Detail the runbook and have two people check while using the console
  4. Turn the work into scripts or CloudFormation and rerun it
Show answer and explanation

Answer: D Turn the work into scripts or CloudFormation and rerun it

Turn repeated work into scripts or IaC so it gives the same result every time. Manual work cannot prevent errors. [Other options] Two-person checks still leave manual errors. Config finds them after the fact. One fixed person creates dependency.

Q2. You want to use the same infrastructure and services as AWS inside your own data center (because some processing cannot send data outside the company). Which service is most suitable?

  1. AWS Direct Connect
  2. AWS Local Zones
  3. AWS Outposts
  4. Amazon CloudFront
Show answer and explanation

Answer: C AWS Outposts

AWS Outposts installs AWS infrastructure and services in the customer’s on-premises facility. Direct Connect is a dedicated connection; CloudFront is content delivery.

Reference links

3.2Understand the AWS global infrastructure

3.2 Define the AWS global infrastructure.

Concept

Regions are geographically separate locations made up of multiple Availability Zones (AZs: one or more data centers with independent power and networking). Edge locations are sites where Amazon CloudFront and others deliver content close to users. There are also Local Zones and Wavelength Zones.

How it works / steps

  1. Choose Regions by data residency regulations, distance to users, price and service availability
  2. For high availability, spread resources across multiple AZs
  3. For disaster recovery, use multiple Regions
  4. Deliver to users worldwide using CloudFront edge locations

Workplace example

Domestic customer data is stored in the Tokyo Region, web servers are spread across two AZs, and disaster recovery backups are kept in the Osaka Region.

Common confusions and exam tips

Mini quiz

Q1. Which is the most appropriate reason to use multiple Regions?

  1. Because multiple AZs cannot be used within one Region
  2. Business continuity against large disasters and low latency for distant users
  3. Because IAM users and roles cannot be created unless you use separate Regions
  4. Because adding Regions makes all charges free
Show answer and explanation

Answer: B Business continuity against large disasters and low latency for distant users

Multiple Regions are used for disaster recovery and business continuity, low latency for distant users and data location requirements. Multiple AZs can be used within one Region.

Q2. Which statement about Availability Zones (AZs) is incorrect?

  1. AZs in the same Region are connected by low-latency networking
  2. Using multiple AZs enables highly available configurations
  3. AZs are designed not to share single points of failure
  4. All AZs in the same Region are located in the same building
Show answer and explanation

Answer: D All AZs in the same Region are located in the same building

AZs are in physically separate locations, not the same building. They are independent and connected with low latency.

Reference links

3.3Identify AWS compute services

3.3 Identify AWS compute services.

Concept

Amazon EC2 provides virtual servers (instance types: general purpose, compute optimized, memory optimized, storage optimized, accelerated computing); AWS Lambda runs code serverlessly; containers use Amazon ECS or Amazon EKS (Kubernetes) with AWS Fargate removing server management; simple web apps can use AWS Elastic Beanstalk or Amazon Lightsail. EC2 Auto Scaling adjusts instance counts automatically, and Elastic Load Balancing distributes load.

How it works / steps

  1. Choose EC2 if you want to manage down to the OS, Lambda for short event processing
  2. For containers, use ECS or EKS; to eliminate server management, use Fargate
  3. Scale instance counts with load using Auto Scaling
  4. Distribute traffic across multiple instances with ELB

Workplace example

When an image is uploaded, Lambda creates a thumbnail, and the web app runs on EC2 in an Auto Scaling group behind ELB.

Common confusions and exam tips

Mini quiz

Q1. Which best describes the role of Amazon EC2 Auto Scaling?

  1. Distribute traffic among multiple instances
  2. Automatically replicate instance data to another Region
  3. Automatically increase and decrease the number of instances according to demand
  4. Automatically apply patches to instance operating systems
Show answer and explanation

Answer: C Automatically increase and decrease the number of instances according to demand

Auto Scaling adjusts the number of instances automatically with demand, providing elasticity. Distributing traffic is the role of Elastic Load Balancing.

Q2. You now run several web servers and want to route incoming traffic only to healthy ones. Which service fits best?

  1. Elastic Load Balancing
  2. Amazon Route 53
  3. Amazon EC2 Auto Scaling
  4. AWS Global Accelerator
Show answer and explanation

Answer: A Elastic Load Balancing

ELB distributes incoming traffic across multiple targets and removes unhealthy targets with health checks. Adjusting counts is Auto Scaling; name resolution is Route 53. [Other options] Auto Scaling adjusts counts. Route 53 is DNS with health checks but not per-connection balancing. Global Accelerator optimizes global routing; in-Region balancing uses ELB.

Reference links

3.4Identify AWS database services

3.4 Identify AWS database services.

Concept

Relational: Amazon RDS (managed MySQL, PostgreSQL and others) and Amazon Aurora (a high-performance AWS engine). NoSQL key-value: Amazon DynamoDB (serverless, millisecond responses). In-memory cache: Amazon ElastiCache. Data warehouse: Amazon Redshift. Graph: Neptune. Document: DocumentDB. Migrate databases with AWS DMS.

How it works / steps

  1. Check the data shape (tables, key-value, documents, graphs) and processing type (transactions, analytics)
  2. For transactional relational workloads, choose RDS or Aurora
  3. For key-value workloads with heavy, low-latency access, choose DynamoDB
  4. Use Redshift for analytics and ElastiCache to speed up reads

Workplace example

An on-premises Oracle database is migrated with DMS to Aurora PostgreSQL; session data goes to ElastiCache and sales analytics to Redshift.

Common confusions and exam tips

Mini quiz

Q1. You are comparing Aurora as a relational database candidate. Which describes Amazon Aurora correctly?

  1. Compatible with MySQL and PostgreSQL, replicating data across multiple AZs
  2. Compatible with PostgreSQL, an appliance installable on-premises
  3. A serverless key-value database you can query with SQL
  4. Compatible with MySQL, storing data in a single AZ
Show answer and explanation

Answer: A Compatible with MySQL and PostgreSQL, replicating data across multiple AZs

Aurora is a MySQL- and PostgreSQL-compatible relational database offered as part of RDS. It replicates data across multiple AZs for durability. [Other options] Aurora replicates data across AZs. It is not an on-premises appliance. Key-value is DynamoDB.

Q2. You want to search complex relationships among people, departments and business partners quickly by traversing the relationships. Which database is most suitable?

  1. Amazon RDS for MySQL
  2. Amazon DocumentDB
  3. Amazon Neptune
  4. Amazon ElastiCache
Show answer and explanation

Answer: C Amazon Neptune

Neptune is a graph database suited to queries that traverse relationships between entities. DocumentDB is for JSON documents; ElastiCache is a cache.

Reference links

3.5Identify AWS network services

3.5 Identify AWS network services.

Concept

Amazon VPC is a logically isolated network within AWS, made up of subnets (public and private), route tables, internet gateways and NAT gateways. Amazon Route 53 is DNS, Amazon CloudFront is content delivery (CDN), and AWS Global Accelerator speeds up traffic over the AWS network.

How it works / steps

  1. Create a VPC, place public web servers in public subnets and databases in private subnets
  2. Public subnets have a route to the internet gateway
  3. Outbound traffic from private subnets uses a NAT gateway
  4. Use Route 53 for domain names and CloudFront for delivering static content

Workplace example

Company website images are placed in S3 and delivered with CloudFront, the domain is managed in Route 53, and the DB sits in a private subnet with no direct external access.

Common confusions and exam tips

Mini quiz

Q1. You want a logically isolated network on AWS where you define your own IP range, subnets and route tables. Which service do you use?

  1. AWS Transit Gateway
  2. AWS Direct Connect
  3. Amazon Route 53
  4. Amazon VPC
Show answer and explanation

Answer: D Amazon VPC

A VPC is a logically isolated virtual network on AWS where the customer defines IP address ranges, subnets, route tables and so on. [Other options] Direct Connect is a dedicated line to on-premises. Transit Gateway is a hub linking many VPCs and networks. Route 53 is DNS.

Q2. You want to route access from users worldwide over the AWS global network to applications in the optimal Region, improving performance and availability. Which service is most suitable?

  1. AWS Direct Connect
  2. AWS Global Accelerator
  3. Amazon VPC peering
  4. AWS Site-to-Site VPN
Show answer and explanation

Answer: B AWS Global Accelerator

Global Accelerator provides static IP addresses and uses the AWS global network to route traffic to the optimal endpoint.

Reference links

3.6Identify AWS storage services

3.6 Identify AWS storage services.

Concept

Amazon S3 is object storage with storage classes by access frequency (Standard, Intelligent-Tiering, Standard-IA, One Zone-IA, Glacier classes). Amazon EBS is block storage for EC2 (within one AZ); Amazon EFS is Linux file storage shared by multiple EC2 instances; FSx provides Windows file servers and more; AWS Storage Gateway connects on-premises and cloud storage; and AWS Backup centrally manages backups.

How it works / steps

  1. Check the type of data (object, block, file)
  2. Choose S3 storage classes by access frequency and transition automatically with lifecycle rules
  3. Use EBS for EC2 OS and DB disks, EFS for sharing among multiple servers
  4. Manage backups together with AWS Backup

Workplace example

Logs are stored in S3 Standard and moved automatically by lifecycle rules to Standard-IA after 30 days and Glacier Deep Archive after one year.

Common confusions and exam tips

Mini quiz

Q1. You want to just upload web application code and have environment setup, load balancing and scaling handled automatically. Which service is most suitable?

  1. AWS Direct Connect
  2. AWS CloudFormation StackSets
  3. Amazon EBS
  4. AWS Elastic Beanstalk
Show answer and explanation

Answer: D AWS Elastic Beanstalk

When you upload code, Elastic Beanstalk automatically handles capacity provisioning, load balancing, Auto Scaling and more.

Q2. Which is the most appropriate use of S3 lifecycle policies?

  1. Automatically reassign storage charges for objects in a bucket to another account after a set period
  2. Move objects older than a set period to cheaper storage classes or delete them
  3. Automatically run a virus scan each time an object is uploaded
  4. Restrict bucket access to specific IP addresses only
Show answer and explanation

Answer: B Move objects older than a set period to cheaper storage classes or delete them

Lifecycle policies automate storage class transitions and deletion based on age, optimizing storage costs.

Reference links

3.7Identify AWS AI/ML services and analytics services

3.7 Identify AWS artificial intelligence and machine learning (AI/ML) services and analytics services.

Concept

AI services: Amazon Rekognition (image and video analysis), Amazon Textract (extracting text and tables from documents), Amazon Comprehend (natural language analysis), Amazon Transcribe (speech transcription), Amazon Polly (speech synthesis), Amazon Translate (translation), Amazon Lex (chatbots), Amazon Bedrock and Amazon Q for generative AI, and Amazon SageMaker AI for ML development. Analytics services: Amazon Athena (query S3 with SQL), AWS Glue (ETL), Amazon QuickSight (BI), Amazon Kinesis (streaming), Amazon EMR (big data) and Amazon OpenSearch Service (search).

How it works / steps

  1. Break the problem into tasks (images, documents, language, speech, generation)
  2. Check whether off-the-shelf AI services suffice; if not, build your own model with SageMaker AI
  3. For data analytics, prepare data with Glue, query with Athena and visualize with QuickSight
  4. Ingest real-time data with Kinesis

Workplace example

Call center recordings are transcribed with Transcribe, sentiment is analyzed with Comprehend, and results are visualized on a QuickSight dashboard.

Common confusions and exam tips

Mini quiz

Q1. A company without machine learning experts wants to adopt AI in its work quickly. Which is the most realistic choice?

  1. Start with fine-tuning a model in Amazon Bedrock
  2. Train a model from scratch in SageMaker AI
  3. Use pretrained AI services such as Textract or Comprehend via API
  4. Run open-source models yourself on EC2 GPU instances
Show answer and explanation

Answer: C Use pretrained AI services such as Textract or Comprehend via API

Pretrained AI services can be used through APIs without ML expertise. If a custom model becomes necessary, consider SageMaker AI. [Other options] Training from scratch or running GPUs needs expertise. Fine-tuning also needs data preparation and evaluation, too heavy as a first step.

Q2. Which pairing of an AI service and what it does is incorrect?

  1. Amazon Textract: extracts text and tables from documents
  2. Amazon Rekognition: analyzes images and videos
  3. Amazon Translate: translates text into another language
  4. Amazon Polly: converts speech to text
Show answer and explanation

Answer: D Amazon Polly: converts speech to text

Polly converts text to speech. Converting speech to text is Transcribe.

Reference links

3.8Identify services from other in-scope AWS service categories

3.8 Identify services from other in-scope AWS service categories.

Concept

Application integration: Amazon SQS (queues), Amazon SNS (notifications and push), Amazon EventBridge (event routing) and AWS Step Functions (workflows). Management and governance: Amazon CloudWatch (monitoring), AWS Systems Manager (operations management) and AWS Trusted Advisor (best practice recommendations). Also in scope: Amazon Connect (contact centers), Amazon WorkSpaces (virtual desktops), AWS IoT Core and Amplify (front-end development).

How it works / steps

  1. Decouple systems with SQS (queues) or SNS (notifications)
  2. Combine multiple steps in order with Step Functions
  3. Set up resource monitoring and alarms in CloudWatch
  4. Check cost, security and performance improvements in Trusted Advisor

Workplace example

Order intake and inventory processing are connected with SQS so order intake doesn’t stop even if inventory processing lags, and anomalies trigger CloudWatch alarms that send email via SNS.

Common confusions and exam tips

Mini quiz

Q1. In the exam guide, which is the best example of a service in the “Customer enablement” category?

  1. Amazon EBS
  2. AWS Support
  3. Amazon SQS
  4. AWS IoT Core
Show answer and explanation

Answer: B AWS Support

The exam guide lists AWS Support as an example of customer enablement services and asks about choosing support that fits the business.

Q2. Choose the incorrect pairing of a messaging/notification service and its use.

  1. Amazon SQS: sends email and SMS to multiple subscribers at once
  2. Amazon EventBridge: routes events by rules
  3. Amazon SES: sends email from applications
  4. Amazon SNS: notifies via publish/subscribe
Show answer and explanation

Answer: A Amazon SQS: sends email and SMS to multiple subscribers at once

SQS is a message queue. Notifying multiple subscribers at once is SNS.

Reference links

Lessons are written by this site based on public information and do not guarantee exam content. Product specifications change, so check the latest information at the reference links.

Notes (unofficial, original questions, no pass guarantee)