Quibank app: now on the App Store — free, works offline. Get the app ↗

21 Technology: Deployment & Global Infrastructure Practice Questions & Answers

Every Technology: Deployment & Global Infrastructure practice question from the AWS Cloud Practitioner Practice Test, with the correct answer and a short explanation.

Start practice test →
  1. 1. A finance analyst with no programming experience needs to check, just this once, which Amazon EC2 instances are running in one Region. Which access method fits best?

    • A.AWS CloudFormation, a template service suited to deploying full stacks
    • B.AWS CLI, a command-line tool suited to commands saved in shell scripts
    • C.AWS Management Console, a browser interface suited to one-time visual tasks✓ Answer
    • D.AWS SDKs, code libraries suited to calling AWS from inside an application

    The AWS Management Console is a browser-based graphical interface that needs no code, which suits a one-off look by a non-technical person. The CLI, SDKs and CloudFormation all add scripting, coding or template work that this single check does not justify.

    Ref QB-YJ768NSource: AWS Certified Cloud Practitioner Exam Guide, Cloud Technology and Services domain: deciding between programmatic access, the console and infrastructure as code; AWS Management Console documentationReport a problem with this question

  2. 2. A development team is building an order-processing application. As part of its normal logic, the application must save each uploaded invoice to Amazon S3. Which access method should the team use?

    • A.The AWS Management Console, so staff upload each invoice through a browser
    • B.The AWS CLI, so an operator types each upload into a terminal as it arrives
    • C.An AWS SDK, so the application's own code calls the service in its language✓ Answer
    • D.An AWS CloudFormation template, so each invoice is declared as a stack resource

    SDKs let application code call AWS services directly in a programming language, which is the cue when the work happens inside an application. Manual methods cannot run as part of the program's logic, and CloudFormation provisions infrastructure rather than handling application data.

    Ref QB-E3PNFKSource: AWS Certified Cloud Practitioner Exam Guide, programmatic access; Tools to Build on AWS (AWS SDKs) documentationReport a problem with this question

  3. 3. A manager asks what the AWS Management Console, the AWS CLI and the AWS SDKs have in common behind the scenes. Which answer is accurate?

    • A.All of them pass each request through an edge location to be cached
    • B.All of them relay each request through a server in the customer's VPC
    • C.All of them ultimately send their requests to the public AWS service APIs✓ Answer
    • D.All of them first turn each request into an AWS CloudFormation stack

    Every way of working with AWS is a layer over the service APIs: the console, the CLI and SDKs all end up making API requests. No stack, customer-run relay server or edge cache is needed for those calls.

    Ref QB-XJUT2MSource: AWS Command Line Interface User Guide, description of direct access to the public AWS service APIs; AWS Certified Cloud Practitioner Exam Guide, ways to access AWS servicesReport a problem with this question

  4. 4. A company must create identical test, staging and production environments in three Regions, and auditors want every infrastructure change tracked. Which approach best meets these needs?

    • A.Build each environment in the AWS Management Console from a detailed written checklist
    • B.Build one environment in one Region and rely on AWS to copy it to the other Regions
    • C.Run AWS Systems Manager commands on existing instances to produce each environment
    • D.Describe the environment in an AWS CloudFormation template kept in version control✓ Answer

    A CloudFormation template provisions the same resources consistently across Regions and accounts, and storing it in version control records each change. Manual console builds drift and invite errors, AWS does not copy environments between Regions on its own, and Systems Manager operates resources that already exist.

    Ref QB-H3RNYPSource: AWS CloudFormation User Guide, What is AWS CloudFormation (replicating infrastructure and tracking changes)Report a problem with this question

  5. 5. A team used one AWS CloudFormation template to create the network, servers and database for a temporary project. The project has ended and everything it created should go away. What is the simplest correct action?

    • A.Delete the stack, since all its resources are managed together as a single unit✓ Answer
    • B.Delete the template file, since the resources exist only while the file does
    • C.Stop the stack's servers, since stopped resources are removed automatically
    • D.Delete each resource by hand, since a stack only records what was created

    CloudFormation manages the resources it deployed from a template as one stack, so deleting the stack deletes those resources. The template is only a text file describing resources: removing it changes nothing that is already running.

    Ref QB-2KWTJQSource: AWS CloudFormation User Guide, concepts: templates and stacksReport a problem with this question

  6. 6. A small team has a finished web application and wants to upload the code and let AWS handle capacity, health monitoring and scaling. Which service meets this need with the least operational overhead?

    • A.AWS Service Catalog, which hosts uploaded code on capacity managed by AWS
    • B.Amazon EC2, which monitors and scales an uploaded application on its own
    • C.AWS Elastic Beanstalk, which deploys uploaded code and manages provisioning✓ Answer
    • D.AWS CloudFormation, which turns uploaded code into a fully managed web server

    Elastic Beanstalk takes application code and handles capacity provisioning, health monitoring and scaling for it. CloudFormation provisions whatever a template describes but does not deploy code by itself, EC2 leaves the servers to you, and Service Catalog offers an approved portfolio of products.

    Ref QB-G3DBTNSource: AWS Elastic Beanstalk Developer Guide, What is AWS Elastic BeanstalkReport a problem with this question

  7. 7. An operations team needs to run the same check on hundreds of Amazon EC2 instances that are already running, without rebuilding any of them. Which service is designed for this?

    • A.AWS Elastic Beanstalk, to upload the check as code to each new instance
    • B.AWS Service Catalog, to publish the check as a product for staff to launch
    • C.AWS Systems Manager, to run a command across instances that already exist✓ Answer
    • D.AWS CloudFormation, to redeploy every instance from a template for each check

    Systems Manager is for operating resources that already exist, including running commands on managed instances at scale. The other services create or package resources, which would mean rebuilding or launching something new rather than acting on the running fleet.

    Ref QB-ZJJGDPSource: AWS Systems Manager User Guide, Run CommandReport a problem with this question

  8. 8. A retailer keeps its own data center for its ordering system, runs analytics on AWS, and connects the two with AWS Direct Connect. Which deployment model does this describe?

    • A.Hybrid, because cloud resources are connected to resources kept outside the cloud✓ Answer
    • B.Cloud, because any workload on AWS makes the whole company count as cloud
    • C.IaaS, because renting virtual servers is the deployment model it has chosen
    • D.On-premises, because owning any data center makes the whole company on-premises

    AWS defines hybrid as connecting cloud resources with existing resources that are not in the cloud, most often to extend on-premises infrastructure. IaaS is a service model, a different classification from cloud, hybrid and on-premises deployment models.

    Ref QB-B3MNXFSource: Overview of Amazon Web Services whitepaper, Types of Cloud Computing (cloud computing deployment models)Report a problem with this question

  9. 9. A startup closed its last data center after moving every application to Amazon EC2 instances inside an Amazon VPC. Which deployment model does it now use?

    • A.SaaS, since the company has stopped operating physical hardware of its own
    • B.On-premises, since a VPC is private and so counts as a private data center
    • C.Hybrid, since its applications were migrated in from a former on-site setup
    • D.Cloud, since all parts of its applications now run in AWS after migration✓ Answer

    In a cloud deployment all parts of the application run in the cloud, whether they were built there or migrated. A VPC is an isolated network in AWS, not an on-site private cloud, and because the company still manages its servers this is not SaaS either.

    Ref QB-UJSTHGSource: Overview of Amazon Web Services whitepaper, Types of Cloud Computing (cloud computing deployment models)Report a problem with this question

  10. 10. A factory must process machine data on site with very low latency, but its engineers want to use the same AWS APIs and tools they use in a Region, on infrastructure that AWS manages. Which service fits?

    • A.AWS Storage Gateway, which runs all AWS compute services behind a local cache
    • B.AWS Outposts, which places AWS-managed infrastructure on the customer's site✓ Answer
    • C.AWS Direct Connect, which runs AWS compute at the plant over a private line
    • D.Amazon CloudFront, which runs local data processing at a nearby edge location

    Outposts brings AWS-managed infrastructure to the customer's premises with the same APIs, which suits low-latency or local data processing. Direct Connect is only a network link, CloudFront delivers content, and Storage Gateway connects on-site applications to AWS storage.

    Ref QB-86QBP4Source: AWS Outposts User Guide, What is AWS OutpostsReport a problem with this question

  11. 11. Which statement correctly describes how Regions, Availability Zones and edge locations relate to one another?

    • A.A Region contains multiple edge locations, and each one holds several Availability Zones
    • B.A Region contains multiple isolated Availability Zones, and edge locations sit outside them✓ Answer
    • C.An Availability Zone contains multiple Regions, and edge locations sit inside each Region
    • D.An edge location contains multiple Availability Zones, and Regions sit inside each of those

    A Region is a separate geographic area made up of multiple isolated Availability Zones. Edge locations are separate content-delivery sites that do not belong to any Availability Zone and are not where you launch instances or databases.

    Ref QB-QNH644Source: AWS Global Infrastructure documentation, AWS Regions and Availability Zones; Amazon CloudFront Developer Guide, edge locationsReport a problem with this question

  12. 12. A colleague says that because Amazon S3 bucket names are unique worldwide, the objects in a bucket must be stored all over the world. Which statement is correct?

    • A.The name is unique globally, so objects are copied to every Region on upload
    • B.The name is unique globally, but the objects stay in the Region you choose✓ Answer
    • C.The name is unique in each Region, so objects stay in the zone you choose
    • D.The name is unique in each account, so objects go to the Region nearest you

    A bucket's name must be unique across its AWS partition, yet the bucket itself is created in one Region and its objects stay there unless you configure replication. Global naming does not mean global storage.

    Ref QB-963PU5Source: Amazon S3 User Guide, general purpose buckets overview and bucket naming rulesReport a problem with this question

  13. 13. A new hire is writing an internal glossary for the company's cloud team. Which definition of an Availability Zone is accurate?

    • A.One or more discrete data centers with redundant power and networking in a Region✓ Answer
    • B.A small site outside any Region that keeps cached copies of content for nearby users
    • C.Exactly one data center building whose power and networking serve an entire Region
    • D.A separate geographic area of the world that holds several fully isolated Regions

    AWS defines an Availability Zone as one or more discrete data centers with redundant power, networking and connectivity in a Region, so it is not necessarily a single building. The cached-content site describes an edge location, and the geographic area describes a Region, which contains Availability Zones rather than the reverse.

    Ref QB-RNVG95Source: AWS Global Infrastructure documentation, Availability ZonesReport a problem with this question

  14. 14. A company wants its web application to keep running if a single data center fails. It wants the lowest added cost and complexity that meets this requirement. What should it do?

    • A.Run the application at edge locations, since they keep it alive when a data center fails
    • B.Run instances in multiple Regions, since nothing else protects against a data center failure
    • C.Run one larger instance in one Availability Zone, since extra capacity absorbs any failure
    • D.Run instances in multiple Availability Zones in one Region, which are isolated from each other✓ Answer

    Availability Zones are isolated from one another's failures, so spreading instances across them survives the loss of a data center. Multiple Regions would also work but add cost and complexity the requirement does not need, one large instance remains a single point of failure, and edge locations reduce latency rather than host the application fleet.

    Ref QB-A6EZZ6Source: AWS Fault Isolation Boundaries whitepaper, Availability Zones; AWS Well-Architected Framework, Reliability PillarReport a problem with this question

  15. 15. A risk officer asks why a failure in one Availability Zone is unlikely to take down another Availability Zone in the same Region. Which explanation is accurate?

    • A.Each sits in a different country, so a local outage cannot reach any of the others
    • B.Each has independent power, cooling and networking and is physically separated✓ Answer
    • C.Each keeps a live copy of every customer's instances, which AWS moves on failure
    • D.Each shares one central power and network core that AWS keeps fully redundant

    Availability Zones do not share single points of failure: each has independent, redundant power, cooling and networking, and they are physically separated within the Region. They are in the same Region, and AWS does not copy your instances between them; you must deploy across zones yourself.

    Ref QB-SN8UE6Source: AWS Fault Isolation Boundaries whitepaper, Availability ZonesReport a problem with this question

  16. 16. A web tier runs on one very large Amazon EC2 instance in one Availability Zone. A manager says the size of the instance already makes the tier highly available. Which assessment is correct?

    • A.It is not highly available, because only a second Region can provide availability
    • B.It is highly available, because large instances run on hardware that cannot fail
    • C.It is highly available, because AWS spreads every instance across zones by default
    • D.It is not highly available, because one instance in one zone can fail alone✓ Answer

    A single instance is a single point of failure whatever its size, and AWS does not spread your deployment automatically: the customer must architect across multiple Availability Zones. A second Region is not required, because multi-AZ is the standard way to reach high availability.

    Ref QB-B6SB78Source: AWS Well-Architected Framework, Reliability Pillar (deploy the workload to multiple locations); Amazon EC2 User Guide, Regions and ZonesReport a problem with this question

  17. 17. A national law requires a company's customer records to remain inside the country where it operates. AWS has a Region in that country. Which choice addresses the requirement?

    • A.Store the data in the Region in that country, because the Region sets where data resides✓ Answer
    • B.Store the data in a second Availability Zone, because the zone sets which country holds it
    • C.Store the data at an edge location in that country, because edge sites are primary storage
    • D.Store the data in the cheapest Region, because AWS meets every local law in all Regions

    Data residency is decided at the Region level, since you choose the Region and your data stays there unless you move it. Availability Zones are subdivisions inside a Region, edge locations hold cached copies rather than primary data, and meeting local law is the customer's responsibility, which outweighs price.

    Ref QB-TNK6M7Source: AWS Architecture Blog, What to consider when selecting a Region for your workloads; AWS Certified Cloud Practitioner Exam Guide, use of multiple RegionsReport a problem with this question

  18. 18. A company keeps its backups in an Amazon S3 bucket in one Region and assumes a copy automatically exists in a second Region for disaster recovery. What should it do to have that second copy?

    • A.Set up Amazon CloudFront, since edge caches keep a lasting copy of each object
    • B.Keep the default setup, since S3 copies each new object to a second Region anyway
    • C.Set up AWS Auto Scaling, since it moves stored data to a healthy Region on failure
    • D.Set up S3 Cross-Region Replication, since Regions copy nothing unless you opt in✓ Answer

    Regions are isolated, and AWS does not replicate your resources across them unless you ask it to, so the company must opt in with a feature such as S3 Cross-Region Replication. Edge caches hold temporary copies for delivery, and Auto Scaling adjusts capacity rather than moving data.

    Ref QB-C65PC9Source: Amazon S3 User Guide, Replicating objects; AWS Global Infrastructure documentation, RegionsReport a problem with this question

  19. 19. A company is reviewing four availability requirements for a workload. Which requirement most clearly justifies the added cost of deploying to a second Region?

    • A.The workload must keep running if one data center in its Region loses power
    • B.The workload must keep running if a large disaster affects an entire Region✓ Answer
    • C.The workload must handle a daily traffic peak from users in a single nearby city
    • D.The workload must keep running if one Amazon EC2 instance fails during the day

    Disaster recovery and business continuity across a Regional event are named reasons for using multiple Regions. A data center or instance failure is handled within one Region using multiple Availability Zones, and a local traffic peak is a capacity matter, not a geographic one.

    Ref QB-UNXGS8Source: AWS Certified Cloud Practitioner Exam Guide, when to use multiple Regions; Disaster Recovery of Workloads on AWS whitepaperReport a problem with this question

  20. 20. A media site stores images in an Amazon S3 bucket in one Region, and viewers on other continents complain of slow page loads. Which service most directly reduces this delay for images that rarely change?

    • A.Amazon CloudFront, to serve cached copies from edge locations near the viewers✓ Answer
    • B.Amazon Route 53, to keep copies of images at edge locations for each viewer
    • C.AWS Global Accelerator, to cache the images behind static anycast addresses
    • D.AWS Auto Scaling, to add instances in the countries where the viewers live

    CloudFront caches content at edge locations, so most viewers are served from a nearby point of presence instead of the origin. Global Accelerator improves network paths without a content caching role, Route 53 answers DNS queries, and Auto Scaling changes capacity rather than geographic placement.

    Ref QB-DWNTUXSource: Amazon CloudFront Developer Guide, How CloudFront delivers contentReport a problem with this question

  21. 21. A multiplayer game sends real-time traffic over UDP to servers in two Regions. The publisher wants fixed IP addresses for players to connect to and fast rerouting around unhealthy endpoints. Which service best fits?

    • A.AWS Global Accelerator, which offers static anycast IPs and routes TCP and UDP traffic✓ Answer
    • B.Amazon CloudFront, which caches each UDP game packet at an edge site for later players
    • C.AWS Direct Connect, which links each player's home to AWS over a private dedicated line
    • D.Amazon Route 53, which carries the UDP game traffic itself across the AWS backbone

    Global Accelerator provides static anycast IP addresses, carries TCP and UDP traffic over the AWS network, and shifts traffic away from unhealthy endpoints quickly, which suits non-HTTP uses such as gaming. CloudFront is for cacheable and dynamic HTTP content, Route 53 answers DNS queries without carrying the traffic, and Direct Connect links a customer's network to AWS.

    Ref QB-WDUYEYSource: AWS Global Accelerator FAQs; AWS Global Accelerator Developer GuideReport a problem with this question

Other topics in this exam

Practice questions built from the domains and task statements in the published AWS exam guide (CLF-C02). AWS and AWS Certified Cloud Practitioner are trademarks of Amazon.com, Inc. or its affiliates; this site is not affiliated with or endorsed by Amazon Web Services, and these are not real exam questions. Fees, scoring, exam languages, Support plans, service names and prices change over time — confirm the current details on the AWS website before you book. The real exam also has multiple-response questions, which this single-answer practice does not reproduce. AWS exam guide (CLF-C02) →