Study guide · Domain 3 · 34% of the exam
Cloud Technology and Services: the AWS Certified Cloud Practitioner objectives in plain English
Domain 3 is the largest part of the AWS Certified Cloud Practitioner exam from Amazon Web Services (AWS), at 34 percent of the exam. It tours what AWS offers: how you deploy and operate in the cloud, how the global infrastructure is built, and which service fits which job across compute, databases, networking, storage, analytics, artificial intelligence, and supporting services. You need to recognize each service from a short scenario, know what problem it solves, and avoid picking a similar sounding service that solves a different problem. Help desk experience helps, because many services are cloud versions of tools you already support.
What the objectives cover
3.1 Define methods of deploying and operating in the AWS Cloud
Match the method to who is doing the work. The AWS Management Console is the browser based graphical interface where you sign in and click through screens to view, create, and manage resources. The AWS Command Line Interface, or CLI, lets you manage services by typing text commands in a terminal, which suits repeatable steps and scripting. AWS software development kits, or SDKs, let application code call AWS services directly from languages such as Python or Java, so programs can call services as they run. Infrastructure as code means defining resources in a template or code so a tool builds them the same way every time. AWS CloudFormation does this for AWS: you write a template describing the resources you want, and CloudFormation creates, updates, and deletes that set as one stack.
3.2 Define the AWS global infrastructure
A Region is a geographic area, such as Northern Virginia or Ireland, made up of multiple isolated locations called Availability Zones. Each Availability Zone is one or more physical data centers with its own power, cooling, and networking inside a Region. Running work across more than one Zone protects it if a single data center fails. Edge locations are a separate, larger set of sites that cache content closer to users. Amazon CloudFront is the content delivery network that uses them, so users download files from a nearby site instead of a distant Region. Choose a Region on four factors. Compliance may require data to stay in a set area. Latency improves when the Region is close to users. Price can differ between Regions. Service availability matters because not every service is offered in every Region.
3.3 Identify AWS compute services
Compute is processing power that runs code, in a different shape for each need. Amazon EC2 provides resizable virtual servers that you configure much like physical servers. AWS Lambda runs code in response to events without provisioning servers, paying only for time it runs. AWS Elastic Beanstalk lets you upload an application while the service handles deployment details such as servers and load balancing. EC2 Auto Scaling automatically adds or removes EC2 instances to match demand. Elastic Load Balancing distributes incoming traffic across multiple targets so no single target is overloaded. Amazon ECS runs and manages Docker containers. Amazon EKS runs Kubernetes containers as a managed service. AWS Fargate is a serverless compute engine for containers, so you manage no underlying servers. AWS Lightsail offers simple virtual servers at a low, predictable price for small projects. AWS Batch plans and runs batch computing jobs, such as processing large sets of files, at scale. AWS Outposts extends AWS infrastructure to your own on premises data center.
3.4 Identify AWS database services
Database questions usually turn on the shape of the data. Amazon RDS is a managed relational service for engines such as MySQL, PostgreSQL, and SQL Server, with AWS handling backups and patching. Amazon Aurora is a relational database built by AWS, compatible with MySQL and PostgreSQL, designed for high performance and availability. Amazon DynamoDB is a serverless NoSQL key value database built for fast performance at large scale. Amazon Redshift is a data warehouse for analytics, built to run complex queries across very large data sets. Amazon ElastiCache adds a fast in memory cache using Redis or Memcached, so frequent reads come from memory instead of the main database. Amazon Neptune is a graph database for data where relationships matter most, such as fraud patterns. Amazon DocumentDB is a document database compatible with MongoDB workloads, storing flexible documents rather than fixed tables. AWS Database Migration Service, or DMS, moves databases to AWS or between engines with limited downtime.
3.5 Identify AWS network services
Networking controls how resources connect to each other, to your office, and to the internet. Amazon VPC, or Virtual Private Cloud, is your isolated section of AWS where you define an address range and launch resources in a network you control. A subnet is a smaller range of addresses inside a VPC, tied to one Availability Zone. Public subnets hold resources that need internet access, while private subnets hold internal systems such as databases. A security group acts as a virtual firewall, with rules allowing specific traffic by port and source. Amazon Route 53 is the Domain Name System service: it registers domains and routes users by translating names into addresses. AWS Direct Connect is a dedicated private connection from your data center or office to AWS, giving more consistent performance than the public internet. An AWS VPN creates an encrypted connection over the internet between your network and your VPC.
3.6 Identify AWS storage services
Storage splits by how data is accessed: as objects, as a disk, or as shared files. Amazon S3 stores data as objects in buckets and suits backups, static website files, images, and logs. Its classes trade access speed for cost. S3 Standard suits data used often. S3 Standard Infrequent Access and S3 One Zone Infrequent Access cost less for data used rarely but still needed quickly. The archive tiers, S3 Glacier Instant Retrieval, S3 Glacier Flexible Retrieval, and S3 Glacier Deep Archive, cost the least and suit long term archives, with retrieval from instant to hours for the deepest tier. Amazon EBS provides block storage volumes attached to a single EC2 instance, like a hard drive for that server, suiting boot volumes and single instance databases. Amazon EFS provides a shared file system that many instances can use at once, suiting shared folders. AWS Storage Gateway connects on premises applications to AWS storage for backup and hybrid access.
3.7 Identify AWS AI/ML services and analytics services
These services provide ready made intelligence or help you work with large amounts of data. Amazon SageMaker builds, trains, and deploys machine learning models. Amazon Rekognition analyzes images and video to identify objects, people, and text. Amazon Polly turns text into lifelike speech. Amazon Lex builds conversational chatbots and voice bots. Amazon Comprehend finds sentiment, topics, and key phrases in text. Amazon Translate translates text between languages. For analytics, Amazon Athena runs SQL queries directly on data in S3 with no server to manage. Amazon EMR runs big data frameworks such as Apache Hadoop and Apache Spark on clusters. AWS Glue is a serverless service that discovers, prepares, and moves data between stores, often called extract, transform, and load work. Amazon Kinesis handles real time streaming data, such as logs and sensor readings, as it arrives. Amazon QuickSight builds business intelligence dashboards and reports.
3.8 Identify services from other in-scope categories
The remaining services fall into clear groups. For messaging and workflow, Amazon SNS is a publish and subscribe service that sends notifications to email, text, and other subscribers. Amazon SQS is a message queue that holds messages between application parts, so one part can slow down without losing work. Amazon EventBridge is a serverless event bus that connects applications using events. AWS Step Functions coordinates multi step workflows as a visual state machine, running steps in order with retries. For contact and email, Amazon Connect is a cloud contact center for phone and chat support, and Amazon SES sends bulk and transactional email. For monitoring and management, Amazon CloudWatch collects metrics, logs, and alarms. AWS CloudTrail records API activity for auditing who did what. AWS Config records how resource configurations change and checks them against rules. AWS Systems Manager provides tools to manage resources, apply patches, and automate routine tasks. AWS Trusted Advisor recommends fixes for cost, performance, security, and limits. Amazon API Gateway creates and manages APIs for reaching back end services, and AWS Amplify helps build and host web and mobile applications.
Worked example
A help desk technician wants to view and manage AWS resources by clicking through a web based interface in a browser, without installing tools or writing code. Which method should the technician use?
The AWS Management Console is a browser based graphical interface for viewing and managing AWS services and resources.
Why the other options are wrong
- BAWS Command Line Interface: The AWS Command Line Interface uses text commands in a terminal, not clicking in a browser.
- CAWS SDKs: AWS SDKs let application code call AWS services, so they require programming.
- DAWS CloudFormation: AWS CloudFormation provisions resources from a template and is not a general browsing interface.
Acronyms and terms in this domain
| Term | What it means |
|---|---|
| EC2 | Elastic Compute Cloud, resizable virtual servers in AWS |
| VPC | Virtual Private Cloud, your isolated network inside AWS |
| S3 | Simple Storage Service, object storage in buckets for files and backups |
| EBS | Elastic Block Store, block volumes attached to EC2 instances |
| EFS | Elastic File System, a shared file system many instances can access at once |
| RDS | Relational Database Service, managed relational databases |
| DNS (Route 53) | Domain Name System, translates domain names into addresses; in AWS this is Route 53 |
| AZ | Availability Zone, one or more data centers inside a Region |
| CDN | Content Delivery Network, caches content at edge locations near users, as CloudFront does |
| IaC | Infrastructure as Code, defining resources in templates or code, for example with CloudFormation |
| SQL vs NoSQL | SQL databases store related data in tables, while NoSQL databases store flexible records such as key value pairs |
| Serverless | A model where the provider runs and scales servers, as with Lambda, and you pay for actual use |
Where the help desk meets this domain
Much of this domain is your daily work with different labels. When you remote into a server to restart a service or check disk space, you are doing the on premises version of what EC2 provides as a virtual server, and the Console is the portal you click through instead of walking to a rack. Shared drives map to the storage choices: a folder only one server uses behaves like an EBS volume, a share a whole team opens at once behaves like EFS, and a web link to a file behaves more like S3 object storage. Archived tickets and old recordings are the rarely touched data that archive classes were built for, and outage dashboards are the same job CloudWatch supports for live health and CloudTrail supports for a record of who changed what.
Support work also explains why the service list is so long. A chat window on a company site, a phone queue, and the email or text that follows a ticket are real help desk tools, and in AWS terms they point to Lex for conversation, Connect for the contact center, and SES or SNS for notifications. When users in another country say a site is slow, you are hearing the Region and latency problem from objective 3.2. When a manager asks for a report on ticket volumes, you are describing analytics work that Athena, Glue, and QuickSight support. For each ticket type you handle, ask which service does that job at scale, and the list becomes familiar tasks with new names.
Traps to watch for
Storage is the most common trap because EBS, EFS, and S3 all store data in different ways. EBS is block storage attached to one EC2 instance, so a boot volume or a disk for a single server points to EBS, even if the word file appears in the stem. EFS is a shared file system, so many instances needing the same files at once points to EFS. S3 is object storage reached over the web, so buckets, static website hosting, backups, and storing images or logs point to S3, not to a disk or a share. The Glacier names are S3 storage classes for archives, not a disk to attach to a server. If a choice offers Glacier for data users need instantly every day, or EBS for files shared by a fleet of servers, the access pattern tells you it is wrong.
Geography wording creates the other major trap. A Region is the geographic area, and an Availability Zone is a data center location inside that Region, so using two Zones in one Region protects against a data center failure but not against loss of the whole Region. Questions often swap the terms, offer an edge location as if it were a place to run databases, or suggest picking the cheapest Region without checking compliance, latency, and service availability. Read the failure the question wants to survive, then pick the matching level. A similar swap happens with deployment tools: clicking in a browser means the Console, typing commands means the CLI, code inside an application means an SDK, and a template that builds repeatable resources means CloudFormation.
Now try it. The 5 free sample questions for this domain are in practice mode, with full explanations.
Practice Domain 3 questions