Architecture questions rarely test one service alone. They describe a workload and ask which combination of services meets its availability, security, cost or performance goal. Each section here draws one common design end to end, so you can see where every service sits and why.
A web application on AWS
Cloud Practitioner questions often describe a simple website and ask which service does which job: where the files live, what spreads the traffic, what keeps it running when a data center fails and who is responsible for what. This design puts all of those pieces on one page.
How it works
- Route 53 answers DNS, and CloudFront serves users from edge locations, with AWS WAF and Shield Standard in front.
- Static files live in Amazon S3; other requests go through the internet gateway to a load balancer.
- EC2 instances in two Availability Zones run the application, and Auto Scaling adds or replaces them.
- Amazon RDS stores the data, with a standby in the second Availability Zone.
- Inspector scans the instances for vulnerabilities, and GuardDuty looks for threats in the logs.
| Part of the design | AWS is responsible for | The customer is responsible for |
|---|---|---|
| EC2 instances | Hardware, network, physical security | Guest OS patches, software, security groups |
| RDS database | Hardware, OS and database patching | Data, access, network settings |
| S3 bucket | Storage infrastructure | Data and who can access it |
| Regions and AZs | The global infrastructure | Choosing where to run |
| Need | Service |
|---|---|
| Domain name and DNS | Route 53 |
| Fast delivery worldwide | CloudFront |
| Filter web attacks | AWS WAF |
| DDoS protection | Shield Standard (automatic), Shield Advanced (paid) |
| Spread traffic | Elastic Load Balancing |
| Match capacity to demand | EC2 Auto Scaling |
| Managed relational database | Amazon RDS |
| Find vulnerabilities | Amazon Inspector |
| Detect threats | Amazon GuardDuty |
Exam cues
- If the question says high availability, think multiple Availability Zones.
- If the question says elasticity, think Auto Scaling.
- If the question says distribute traffic, think a load balancer.
- If the question says block SQL injection or filter web requests, think AWS WAF.
- If the question says the customer’s security responsibility on EC2, think the guest OS, software and security groups.
Remember
- Region: a geographic area with several Availability Zones. Edge location: CloudFront cache near users.
- Security groups are stateful; network ACLs are stateless.
- Public subnet: route to an internet gateway.
Trap: Running in one Availability Zone is not highly available, even with several instances. A zone failure takes them all down.
Users look up the site's name in Route 53, the AWS DNS service.
Requests reach the nearest CloudFront edge location. AWS WAF filters them, and Shield Standard protects against common DDoS attacks.
Images and files come from an S3 bucket, cached at the edge.
Other requests go through the VPC's internet gateway to the load balancer in the public subnets.
The load balancer spreads requests over EC2 instances in two Availability Zones. An Auto Scaling group adds or replaces instances as needed.
The instances use an RDS database. RDS keeps a synchronous standby copy in the second Availability Zone.
Amazon Inspector scans the instances for software vulnerabilities.
GuardDuty analyzes the account's logs, such as VPC flow logs and DNS logs, to detect threats.
If Availability Zone A fails, RDS switches to the standby, and the Auto Scaling group launches new instances in the healthy zone.
Detalhes das caixas (4)
- CloudFront
Shield Standard protects it against common DDoS attacks at no extra cost.
- S3 bucket
Private. CloudFront reaches it through origin access control.
- Load balancer
Its security group allows HTTPS from the internet; the instances' security group allows traffic only from the load balancer.
- EC2 instances
Part of an Auto Scaling group across both Availability Zones.
Da documentação da AWS (32)
- Security and compliance are a shared responsibility between AWS and the customer, commonly described as security of the cloud versus security in the cloud.
- Security in the cloud is the customer's part; it is determined by the services used, and also depends on the sensitivity of the data, the company's requirements and applicable laws and regulations.
- For Amazon EC2, an IaaS service, the customer manages the guest operating system including updates and security patches, any software installed on the instances, and the security group firewall configuration.
- Amazon RDS automatically schedules required operating system and database patching for patches related to security and instance reliability.
- AWS is responsible for the global infrastructure that runs AWS services: the Regions, Availability Zones and edge locations.
- A Region is a geographical location where AWS clusters data centers; each Region is made of multiple independent, physically separate Availability Zones.
- An Availability Zone is one or more discrete data centers with independent and redundant power, networking and connectivity in a Region.
- An application that uses multiple Availability Zones is better protected from issues that affect one Availability Zone and, if architected correctly, stays up when one fails.
- DNS routes a CloudFront request to the edge location (point of presence) that can best serve it, typically the nearest in latency, which returns cached content or fetches it from the origin.
- AWS WAF is a web application firewall that watches HTTP(S) requests to resources such as CloudFront distributions and API Gateway APIs and allows or blocks them by rules.
- AWS gives automatic protection against common DDoS attacks such as UDP reflection and SYN floods at no extra cost.
- Amazon GuardDuty is a threat detection service that continuously analyzes AWS data sources and logs, using threat intelligence and machine learning, to find malicious activity such as compromised credentials.
- GuardDuty automatically ingests CloudTrail management events, EC2 VPC flow logs, and DNS logs as foundational data sources.
- Amazon Inspector automatically discovers workloads such as EC2 instances and continually scans them for software vulnerabilities and unintended network exposure.
- Amazon EC2 Auto Scaling keeps the right number of EC2 instances for the load, in collections called Auto Scaling groups with a minimum, desired and maximum size; scaling policies change the count within those limits.
- An Auto Scaling group launches replacement instances after unexpected termination to maintain desired capacity.
- Elastic Load Balancing spreads incoming traffic across targets such as EC2 instances, containers and IP addresses in one or more Availability Zones, and sends traffic only to healthy targets.
- Amazon RDS is a managed service that makes it easier to set up, operate and scale a relational database and manages common database administration tasks.
- Amazon RDS manages backups, software patching, automatic failure detection and recovery, and offers high availability with a synchronous secondary instance and read replicas for read scaling.
- Amazon VPC launches AWS resources in a logically isolated virtual network that you define, closely resembling a traditional data center network.
- A subnet is a range of IP addresses in a VPC and must reside in a single Availability Zone.
- A public subnet has a direct route to an internet gateway; a private subnet doesn't, and its resources need a NAT device to reach the internet.
- An internet gateway is a redundant, highly available VPC component for internet communication. Resources need a public IPv4 or public IPv6 address, an appropriate subnet route, and security controls that allow the traffic.
- A security group acts as a virtual firewall for an instance: only traffic its rules allow reaches it, and it is stateful, so responses are allowed back automatically.
- A network ACL allows or denies inbound and outbound traffic at the subnet level.
- Security groups track connections and automatically allow response traffic for allowed requests regardless of opposite-direction rules.
- Network ACLs do not track connections. Traffic between subnets must pass the source outbound and destination inbound rules; replies must pass both subnet ACLs in the reverse direction.
- Amazon Route 53 is a highly available and scalable DNS web service for domain registration, DNS routing and health checking.
- Amazon S3 is an object storage service for storing and protecting any amount of data for uses such as data lakes, websites, mobile applications, backup and restore, archive and big data analytics.
- Amazon S3 is designed for 99.999999999% (11 nines) durability, and S3 Standard for 99.99% availability.
- Amazon CloudFront speeds up delivery of static and dynamic web content by serving it from a worldwide network of edge locations.
- Origin access control authenticates CloudFront requests to an S3 bucket origin.
Migrating a data center to AWS
Migration questions describe a company leaving its data center and ask which strategy fits each workload and which AWS service moves it. This design shows the common pieces together: servers rehosted, a database replatformed, files transferred, a gateway for files that stay and a private network link.
How it works
- Direct Connect gives a private connection for migration traffic; a Site-to-Site VPN is the backup.
- AWS Transform MGN rehosts servers on EC2 without changing them.
- Schema conversion and AWS DMS replatform the database onto Aurora, keeping it in sync until cutover.
- DataSync copies file shares to S3, and Storage Gateway lets on-premises applications keep using S3 through file shares.
| Strategy | What changes | Example |
|---|---|---|
| Rehost | Nothing | Servers to EC2 with Transform MGN |
| Replatform | Some optimization | Oracle to Aurora with DMS |
| Refactor | The application is re-architected | Most complex and costly |
| Retire, retain, relocate, repurchase | The other four of the 7 Rs | Chosen per application |
| Need | Service |
|---|---|
| Move servers unchanged | AWS Transform MGN |
| Convert a database schema | AWS SCT or DMS Schema Conversion |
| Move database data | AWS DMS |
| Copy files online | AWS DataSync |
| On-premises access to S3 storage | AWS Storage Gateway |
| Private dedicated network link | AWS Direct Connect |
| Connection over the internet, or backup | AWS Site-to-Site VPN |
| Business case for migration | Migration Evaluator |
Exam cues
- If the question says lift and shift, think rehost with AWS Transform MGN.
- If the question says migrate a database with minimal downtime, think AWS DMS with ongoing replication.
- If the question says convert the schema to another engine, think AWS SCT or DMS Schema Conversion.
- If the question says transfer files online, think DataSync.
- If the question says on-premises applications use cloud storage, think Storage Gateway.
- If the question says a private connection that avoids the internet, think Direct Connect.
Remember
- The 7 Rs: retire, retain, rehost, relocate, repurchase, replatform, refactor.
- Schema conversion changes the schema; DMS moves the data.
- Direct Connect does not use the public internet; a VPN does.
Trap: Snowball Edge is no longer offered to new customers. For online transfers, think DataSync.
A Direct Connect connection links the data center to AWS without going over the public internet, so migration traffic has a private path.
Rehost: AWS Transform MGN replicates the servers continuously and launches them as EC2 instances, unchanged.
Replatform: the schema is converted for Aurora, then AWS DMS copies the data and keeps replicating changes until the cutover.
DataSync copies the file shares to an S3 bucket online.
Applications that stay on premises keep writing files through Storage Gateway, which stores them in S3.
If the Direct Connect link is down, a Site-to-Site VPN over the internet keeps the data center connected.
Detalhes das caixas (4)
- Storage Gateway
Gives on-premises applications NFS or SMB shares whose files are stored as objects in S3.
- Site-to-Site VPN
Two tunnels; if one is unavailable, traffic uses the other.
- Transform MGN
Formerly AWS Application Migration Service. Replicates servers at the block level and converts them to run on AWS.
- AWS DMS
Schema conversion (AWS SCT or DMS Schema Conversion) converts the schema first; DMS then moves the data.
Da documentação da AWS (24)
- The seven migration strategies, the 7 Rs, are retire, retain, rehost, relocate, repurchase, replatform, and refactor or re-architect.
- Rehost, also called lift and shift, moves an application to AWS without changing it.
- Replatform, also called lift, tinker and shift, moves an application and adds some optimization, such as moving a SQL Server database to Amazon RDS for SQL Server.
- Refactoring is the most complex and costly migration strategy; for large migrations AWS recommends rehosting, relocating or replatforming first and modernizing afterward.
- AWS Application Migration Service, now named AWS Transform MGN, automates migrating physical, virtual and cloud servers to AWS with minimal downtime, typically with cutover windows of minutes.
- AWS Transform MGN continuously replicates source servers at the block level and converts them to launch on AWS.
- AWS Application Migration Service is the primary service AWS recommends for lift-and-shift migrations.
- Migration Evaluator builds a data-driven business case for AWS, discovers over-provisioned on-premises instances and compares BYOL with License Included options.
- AWS Migration Hub provides a single place to discover servers, plan migrations and track the status of each application migration, whichever migration tool is used.
- AWS Migration Hub closed to new customers on November 7, 2025; AWS Transform offers similar capabilities.
- AWS Database Migration Service (AWS DMS) migrates relational databases, data warehouses, NoSQL databases and other data stores into AWS or between cloud and on-premises setups.
- AWS DMS can do a one-time migration or replicate ongoing changes to keep source and target in sync.
- The AWS Schema Conversion Tool (AWS SCT) converts an existing database schema from one database engine to another, for OLTP or data warehouse schemas.
- DMS Schema Conversion converts the database schema, not the data; data migration uses the data migration features of AWS DMS.
- AWS DataSync is a secure, high-speed service for transferring file or object data to, from and between AWS storage services.
- AWS Snowball Edge is no longer available to new customers; AWS points new customers to DataSync for online transfers, AWS Data Transfer Terminal for physical transfers, and Outposts for edge computing.
- AWS Storage Gateway connects an on-premises software appliance with AWS cloud storage so on-premises applications can use it.
- S3 File Gateway exposes S3 objects as files through Network File System (NFS) or Server Message Block (SMB) shares.
- AWS Site-to-Site VPN connects a VPC to a remote network, such as an on-premises site, which instances in a VPC cannot reach by default.
- Configure both tunnels for redundancy; when one becomes unavailable traffic automatically uses the available tunnel.
- AWS Direct Connect links an internal network to a Direct Connect location over a standard Ethernet fiber-optic cable, without the public internet.
- For identical destination prefixes at a virtual private gateway, BGP routes from Direct Connect take priority over Site-to-Site VPN routes, subject to endpoint health.
- A Site-to-Site VPN can act as a backup path that keeps VPC connectivity if the Direct Connect connection fails.
- Direct Connect does not encrypt traffic in transit by default.
Flashcards
Nada para revisar aqui agora. Escolha outros tópicos ou outro filtro.
·
Todos os 19 flashcards em lista
In this web application, who patches the EC2 instances' operating system, and who patches the RDS database engine?
The customer patches the guest OS on EC2. For RDS, AWS schedules the required OS and database patches. AWS always secures the physical infrastructure.
How does this design keep running when one Availability Zone fails?
It runs in two Availability Zones: the Auto Scaling group replaces lost instances, and RDS switches to its standby in the other zone.
Traffic doubles every evening and drops at night. Which feature adds and removes EC2 instances to match?
Amazon EC2 Auto Scaling, which keeps the number of instances between a minimum and maximum. This is elasticity.
What does the load balancer do in this design?
It spreads incoming requests across healthy targets in several Availability Zones, so no single instance takes all the traffic.
Why are the EC2 instances and the database in private subnets, while the load balancer is in public subnets?
A public subnet has a route to the internet gateway; a private one does not. Only the load balancer needs to be reachable from the internet.
Which works at the instance level and remembers connections: a security group or a network ACL?
A security group: instance level, stateful, allow rules only. A network ACL works at the subnet level and is stateless.
Match each service to its job: AWS WAF, AWS Shield Standard, Amazon GuardDuty, Amazon Inspector.
WAF filters web requests. Shield Standard blocks common DDoS attacks at no extra cost. GuardDuty detects threats in logs. Inspector scans for software vulnerabilities.
Which AWS service gives the application its domain name and DNS routing?
Amazon Route 53.
Where should the application keep user-uploaded images, and how do users far away get them quickly?
In Amazon S3, object storage designed for 11 nines of durability, served through CloudFront edge locations.
Is an edge location the same as an Availability Zone?
No. An Availability Zone is one or more data centers inside a Region. An edge location serves CloudFront content close to users.
A company wants to move hundreds of servers to AWS quickly, without changing the applications. Which strategy and service fit?
Rehost (lift and shift) with AWS Transform MGN, formerly AWS Application Migration Service, which replicates servers and launches them on AWS.
During migration, a self-managed Oracle database moves to a managed engine on AWS with small changes. Which strategy is this?
Replatform (lift, tinker and shift): move and make some optimizations, such as a managed database.
Which AWS tools convert a database schema to a new engine and then copy the data while the source keeps running?
AWS SCT (or DMS Schema Conversion) converts the schema; AWS DMS migrates the data and can keep replicating changes.
Does DMS Schema Conversion also move the rows of the database?
No. It converts only the schema. The data moves with AWS DMS.
Large file shares must be copied to S3 over the network as fast as possible. Which service fits?
AWS DataSync, a secure, high-speed online transfer service.
On-premises applications must keep using file shares, but the files should be stored in Amazon S3. Which service fits?
AWS Storage Gateway with an S3 File Gateway, which offers NFS or SMB shares backed by S3.
What is the difference between AWS Direct Connect and AWS Site-to-Site VPN?
Direct Connect is a dedicated link that does not use the public internet. Site-to-Site VPN connects over the internet and can be the backup path.
Before migrating, management wants a data-driven business case that compares on-premises costs with AWS. Which tool fits?
Migration Evaluator, which builds a business case and finds over-provisioned on-premises servers.
Which migration strategy is usually the most complex and costly?
Refactor (re-architect). For large migrations, AWS recommends starting with rehost, relocate or replatform.