Task 1.2: Describe the benefits of using cloud services

• AAdd VM instances behind a load balancer and spread the work across them

  Incorrect

  Adding instances is horizontal scaling. It helps apps that can run as several copies, which this one can't.

• BResize the VM to a size with more vCPUs and memory

  Correct

  Scaling up, or vertical scaling, adds CPU or RAM to the existing resource, which is how a single-instance app gets more capacity.

• CPlace the VM in an availability set with two fault domains

  Incorrect

  Availability sets protect against hardware failure and maintenance events. They don't give a single VM more CPU.

• DAttach a premium SSD data disk to the VM

  Incorrect

  A faster disk improves storage IOPS and throughput. The bottleneck here is CPU.

Vertical scaling adds capabilities to an existing resource, such as more CPUs or RAM for a VM. Horizontal scaling adds more resources, which only helps an app built to run as several instances.

• AAbout 99.99%, the higher of the two targets

  Incorrect

  A request fails if either part is down, so the whole can't be more available than its weaker part.

• BAbout 99.97%, the average of the two targets

  Incorrect

  Averaging doesn't model dependencies. When both are needed, their availabilities multiply.

• CAbout 99.95%, the lower of the two targets

  Incorrect

  The lower target is an upper bound. The database's own downtime is added on top, which pulls the combined figure below it.

• DAbout 99.94%, the product of the two targets

  Correct

  A composite target is the product of its dependencies: 0.9995 x 0.9999 is roughly 0.9994, slightly below either part alone.

When a workload needs several components at once, its composite availability is the product of theirs, so it's lower than any single component's. That's why chaining services lowers the overall target and why redundancy is added to lift it.

• APredictability

  Incorrect

  Predictability doesn't address keeping an app running when a region goes offline.

• BManageability

  Incorrect

  Manageability is about how resources are managed, not about keeping the app running when a region goes offline.

• CReliability

  Correct

  Reliability is a system's ability to recover from failures and keep working, and the cloud's global spread of regions supports it.

• DGovernance

  Incorrect

  Governance concerns standards for resources, not keeping the app running when a region goes offline.

Microsoft describes reliability as the ability of a system to recover from failures and continue to function. Because the cloud spreads resources across regions worldwide, an app designed to use more than one region keeps running even if one region has a catastrophic event.

• AManaging resources through the Azure portal

  Incorrect

  A web portal is a way of managing in the cloud: it's the interface you use, not something the cloud does for you.

• BRunning Azure CLI commands from a script

  Incorrect

  A command-line interface is one of the ways you manage in the cloud.

• CCalling the Azure REST APIs from an internal tool

  Incorrect

  APIs are another channel for managing in the cloud, alongside the portal, CLI and PowerShell.

• DScaling resources automatically as demand changes

  Correct

  Automatic scaling based on need is listed under management of the cloud.

• EReplacing failing resources automatically after a health check

  Correct

  Monitoring resource health and automatically replacing failing resources is listed under management of the cloud.

Management of the cloud covers what you can have the cloud do with your resources: scale automatically, deploy from templates, monitor health and replace failing resources, and raise alerts. Management in the cloud covers how you do it: a web portal, a command line, APIs or PowerShell.

• APaaS, with patches and maintenance handled for the firm

  Incorrect

  PaaS is the choice Microsoft describes for having patches and maintenance taken care of automatically, not for maximum control of security.

• BIaaS, with the firm managing the OS and its patches

  Correct

  Microsoft notes that IaaS gives maximum control of security, and in exchange you manage the operating systems and installed software, including patches and maintenance.

• CIaaS, with patches and maintenance handled for the firm

  Incorrect

  IaaS gives the most control, but that control comes with managing the operating system and its patches yourself, not with automatic patching.

• DPaaS, with the firm managing the OS and its patches

  Incorrect

  In PaaS the provider maintains the operating system, so the firm neither manages it nor gets the most control of security.

Microsoft describes security as a benefit you can match to your needs. If you want maximum control of security, IaaS gives it to you, but you manage the operating systems and installed software, including patches and maintenance. If you want patches and maintenance taken care of automatically, PaaS or SaaS may be the better strategy.

• AScalability

  Incorrect

  Scalability is adjusting resources to meet demand. Forecasting what the bill will be is a different benefit.

• BCost predictability

  Correct

  Cost predictability is forecasting cloud spend by tracking use in real time and analyzing patterns and trends.

• CPerformance predictability

  Incorrect

  Performance predictability is about having the resources needed for a good customer experience, not forecasting the bill.

• DHigh availability

  Incorrect

  High availability keeps resources available despite disruptions; it says nothing about forecasting cost.

Predictability in the cloud covers performance and cost. Cost predictability means forecasting cloud spend: you track resource use in real time, find patterns and trends, predict future costs and adjust resources, with tools such as the pricing calculator for estimates.

• AManageability, because resources can be deployed from templates

  Incorrect

  Deploying from templates removes manual configuration. It doesn't absorb an attack.

• BPredictability, because the bill can be forecast in advance

  Incorrect

  Knowing the bill ahead of time doesn't protect the service from an attack.

• CSecurity, because cloud providers are well suited to handle DDoS attacks

  Correct

  Microsoft lists this as a security benefit: providers are typically well suited to handle DDoS attacks, making the network more robust.

• DScalability, because a larger VM can take on the attack traffic

  Incorrect

  A bigger VM is still one target, and scaling up doesn't filter attack traffic.

Among the security benefits of the cloud, Microsoft notes that cloud providers are typically well suited to handle distributed denial of service attacks, which makes a network more robust and secure than a single on-premises link.

• AScaling down

  Incorrect

  Scaling down lowers the CPU or RAM of a resource. The store kept the same size and changed the number of instances.

• BScaling up

  Incorrect

  Scaling up adds CPU or RAM to a resource; nothing was made larger here.

• CFailing over

  Incorrect

  Failover moves a workload to another location after a failure; the store only removed capacity it no longer needed.

• DScaling in

  Correct

  Reducing the number of resources, such as VM instances, after demand drops is scaling in, the reverse of scaling out.

Horizontal scaling changes the number of resources: scale out adds instances when demand rises, and scale in removes them when it drops. Vertical scaling changes a resource's capabilities: scale up or down. Scaling in lets the store stop paying for unused capacity.