Task 3.2: Define the AWS global infrastructure.

• AA single data center that shares its power supply with the other zones in the Region

  Incorrect

  Availability Zones have separate, redundant power, and common points of failure such as generators and cooling are not shared between them.

• BA separate AWS Region reserved for disaster recovery

  Incorrect

  Availability Zones are locations inside a Region, not Regions of their own.

• COne or more discrete data centers in a Region with separate, redundant power

  Correct

  That is the definition: one or more discrete data centers in a Region, isolated from the other zones so that they do not fail together.

• DA set of data centers in different countries, connected over the public internet

  Incorrect

  The zones of a Region are close enough for synchronous replication and are connected by dedicated, low-latency metro fiber.

Zones are far enough apart not to share a failure, and close enough to replicate synchronously — which is why spreading an application across two or more of them raises its availability.

• AIn every Availability Zone of the website's Region

  Incorrect

  Availability Zones are data centers inside one Region. Caching the content there would still leave it far from readers on other continents.

• BIn edge locations

  Correct

  CloudFront delivers content through a worldwide network of edge locations and routes each request to the one with the lowest latency, where copies of the files are cached.

• CIn AWS Local Zones

  Incorrect

  Local Zones extend a Region so you can run compute and storage close to a large population center. CloudFront's caches are its edge locations.

• DOnly in the origin's Region, which CloudFront makes faster

  Incorrect

  The origin keeps the definitive copy, but CloudFront's whole point is to cache copies away from it, near the viewers.

Regions and their Availability Zones hold your resources; CloudFront's edge locations hold cached copies of your content close to the people reading it.

• AThe business must keep operating even if the whole Region is disrupted

  Correct

  Regions are isolated from each other, so continuity requirements that cannot be met inside one Region are a common reason to fail over to another.

• BThe application must survive the failure of a single data center

  Incorrect

  Availability Zones already provide this: they do not share power, cooling or other common points of failure, so three zones cover it.

• CCustomers on another continent need lower latency

  Correct

  Deploying in Regions close to users around the world gives them lower latency and a better experience.

• DTraffic must be spread across the application's instances

  Incorrect

  A load balancer distributes traffic across targets in one or more Availability Zones of a Region; no second Region is needed.

• EAWS must copy the data to other Regions automatically

  Incorrect

  Data created in one Region does not exist in any other unless you replicate or copy it yourself, with a service feature or your own tooling.

Several Availability Zones handle the failure of a data center. A second Region is for what one Region cannot give: surviving a Region-wide event, being near users elsewhere, or keeping data in a particular jurisdiction.

• AEach Region has multiple, physically separate Availability Zones

  Correct

  Each Region is made of multiple independent and physically separate Availability Zones within a geographic area; all Regions currently have three or more.

• BEach Availability Zone contains several Regions

  Incorrect

  The nesting is the other way round: Availability Zones are inside a Region.

• CAn edge location is an Availability Zone reserved for Amazon CloudFront

  Incorrect

  Edge locations are a separate worldwide network of data centers that CloudFront uses to cache content; they are not Availability Zones.

• DA Region is a single data center

  Incorrect

  A Region contains several Availability Zones, and each zone is itself one or more discrete data centers.

Region contains Availability Zones; Availability Zones contain data centers; edge locations sit outside this hierarchy, closer to users.

• AMove both instances to a larger instance type

  Incorrect

  A larger instance adds capacity, but both instances would still be in the one zone and fail with it.

• BAdd a third instance in the same Availability Zone

  Incorrect

  More instances in the same zone still share its fate; they all go down if the zone does.

• CRun the instances on a Dedicated Host

  Incorrect

  A Dedicated Host is a single physical server dedicated to you. It concentrates the instances rather than spreading them.

• DUse two Availability Zones and a load balancer

  Correct

  Availability Zones do not share common points of failure, and a load balancer can spread traffic across targets in several zones and route only to healthy ones.

High availability inside a Region comes from spreading across Availability Zones, with a load balancer sending traffic to whatever is healthy.