Chapter 1
The Principles of Cloud Computing
CERTIFICATION OBJECTIVES
1.01 Identifying Cloud Computing Characteristics
1.02 Identifying Cloud Service Models
1.03 Identifying Cloud Deployment Models
1.04 Understanding Cloud Shared Responsibility
Two-Minute Drill
Q&A Self Test
The CompTIA Cloud Essentials+ exam is designed to test whether a candidate has knowledge of how cloud computing solutions address business problems. This chapter introduces you to cloud computing concepts. It defines the characteristics of cloud computing and identifies cloud service models and cloud deployment models. This chapter also helps you to recognize where responsibility lies when deploying different types of cloud services. You finish this chapter by learning about the essential role of hypervisors and virtualization in cloud computing.
CERTIFICATION OBJECTIVE 1.01
Identifying Cloud Computing Characteristics
Cloud computing, from a simplistic view, allows you to run IT solutions on somebody else’s equipment. Technology that has been around for a long time is simply being used in a different way over a network. But is it really that simple? Of course not! If it were, you wouldn’t need to study hard to obtain the CompTIA Cloud Essentials+ certification. As the old saying goes, “The devil is in the details.”
On the Internet, cloud service providers (CSPs) are also application service providers (ASPs) that provide the IT infrastructure and services that allow the use of IT systems over a network. While ASPs are focused on applications, CSPs offer the provisioning of apps and the underlying infrastructure. Cloud customers are also called cloud tenants. The cloud is described as a multitenant environment, much like an apartment building hosts multiple tenants that use the same infrastructure. A CSP keeps the data and configuration settings for each cloud tenant isolated from those of other cloud tenants. But what exactly is “the cloud”? To define this concept, we will examine the common characteristics shared by cloud computing implementations.
You might see some exam questions that simply refer, for example, to a CSP rather than to a cloud service provider. Make sure you are familiar with all acronyms in this book and in the CompTIA Cloud Essentials+ Acronyms list that follows the official exam objectives; terms won’t always be written out for you on the exam.
Elasticity and Scalability
To conceptualize elasticity as a cloud characteristic, think of an elastic rubber band. When you pull it, the rubber band stretches to accommodate the change, and when you release the tension, it shrinks back to its normal size. In the cloud, elasticity is the dynamic provisioning and deprovisioning of resources to meet demand. Examples of this include
Increasing or decreasing the amount of cloud storage
Adding or removing virtual machines (VMs) to support an e-commerce web application
Adding virtual machines to support a busy application is also called scaling out, a form of horizontal scaling. When application requests return to a normal level, the removal of VMs is called scaling in.
In the cloud, an autoscaling configuration for an application could be configured to keep a minimum of two virtual machines running but can grow to six virtual machines if the virtual machines’ central processing unit (CPU) percentage thresholds are violated. You can see an example of this configuration in Figure 1-1.
FIGURE 1-1 Microsoft Azure virtual machine scale set autoscaling
Besides scaling in and out, or horizontal scaling, vertical scaling is another option to increase or decrease performance (and costs). With vertical scaling, scaling up refers to increasing existing virtual machine virtual CPU and RAM settings to improve application performance. Scaling down reduces the number of virtual CPUs and RAM (and cost!). Figure 1-2 shows virtual machine sizing for an existing Microsoft Azure virtual machine.
FIGURE 1-2 Microsoft Azure virtual machine sizing
Self-service
Without cloud computing, if you need to deploy a new virtual machine or database, you probably need to submit a request to the IT department to make it happen. Cloud computing puts deployment and full or limited management control into the hands of the cloud computing user.
Cloud users can deploy and manage cloud services on demand through a graphical user interface (GUI), which normally comes in the form of a web application or a mobile device app, an example of which is shown in Figure 1-3.
FIGURE 1-3 Amazon Web Services (AWS) Console app installation screen
Cloud resources can also be deployed and managed using command-line interface (CLI) tools. Major public CSPs, such as AWS and Microsoft Azure, offer cloud management using Microsoft PowerShell cmdlets, as well as a CLI that can run on Linux and macOS. Developers can also access cloud resources programmatically through application programming interface (API) calls.
API calls made over the Hypertext Transfer Protocol (HTTP) are done using what is called the Representational State Transfer (REST) API. CLI and API access to cloud resources lends itself to automation and repeatability.
Broad Network Access
In the business world, smartphones, tablets, laptops, and desktop computers all can be used to access cloud resources, making broad network access a key characteristic of cloud computing. Additionally, specialty devices in manufacturing, medicine, and other fields and Internet of Things (IoT) devices such as smart cars and baby monitors send data over the Internet. All of these devices can be configured to access cloud services.
Using any type of device to access IT solutions over a network is what broad network access refers to. Depending on the cloud service, a web browser or a specific app might be required on the device. For example, users can access Dropbox cloud storage using a web browser or an app. Security is ensured by using generic apps such as the Microsoft VPN client to establish an encrypted virtual private network (VPN) tunnel between the client device and a cloud-based virtual network.
On the exam, watch out for questions related to the “broad network access” cloud computing characteristic. Broad network access does not necessarily imply accessing cloud services over the Internet. Accessing an organization’s private cloud services internally occurs over a local area network (LAN).
Pay-as-You-Go
Some cloud subscriptions charge a small monthly fee in addition to charges for cloud resources used during the month. Other subscriptions might only charge for usage and nothing else. For organizations, using cloud computing means what was formally an on-premises capital expense (CAPEX) for equipment, software, licensing, and support is now a monthly operating expense (OPEX).
Think of how you pay for water or electricity—it’s based on your consumption; you only pay for what you use—this is what pay-as-you-go is all about. You might also hear this referred to as metered usage.
You can quickly and easily spin up a dozen virtual machines in the cloud. Once you shut down the VMs, you are no longer charged, although there might still be small charges for storing the VM hard disks or public IP addresses you might have assigned to VM network interfaces. Compare this to having to acquire and pay for the hardware and software to accomplish the same task—cloud computing, at least in the short term, is usually less expensive than provisioning the same services yourself on your network.
Availability
CSPs publish service level agreements (SLAs) for each cloud service offering, such as cloud storage, virtual machines, and databases. One provision contained within the standard SLA is a designated level of guaranteed uptime (such as 99.9 percent), which relates to service and data availability. Often, the SLA provides details on cloud computing credits allocated to customers who experience outages that fall within the scope of the SLA.
Another aspect of availability falls on the shoulders of the cloud customer. If you depend on public cloud computing services and your Internet link goes down, you have a problem. Organizations should consider implementing redundant Internet links through different Internet service providers (ISPs) to maximize cloud service and data availability.
Individuals and organizations can also synchronize or back up data to the cloud—this is commonly done by smartphone users. If a user replaces her smartphone, she can simply sign in to her cloud account on the new phone to gain access to pictures, music, contact lists, and so on. A smartphone app might need to be installed to allow the user to first sign in to her cloud account. Figure 1-4 shows synced Google account items on an Android smartphone.
FIGURE 1-4 Google account sync settings
CERTIFICATION OBJECTIVE 1.02
Identifying Cloud Service Models
Humans like to categorize, well, pretty much everything. Cloud service offerings are no exception. Some cloud service models appeal to end users, others to developers, others to server administrators, security professionals, and so on. As a CompTIA Cloud Essentials+ exam candidate, you need to understand these cloud service models. Cloud services model names normally begin with the type of service followed by “as a service,” as you will see in the following text.
Software as a Service (SaaS)
Years ago, before you could run an application such as e-mail, you would first need to install and configure the software on your computer. If you didn’t have a mail server to connect to, you’d also have to acquire the hardware to run the server, then install and configure the e-mail server software—an especially long task if you were installing from floppy disks back in the day.
With cloud computing, accessing hosted software over a network is referred to as Software as a Service (SaaS). Users typically access a SaaS solution through a web browser or through an installed app. With SaaS, users are essentially renting the use of software over a network, although some services are free. SaaS examples include
Gmail
Dropbox
Netflix
Uber
Microsoft Office 365
Infrastructure as a Service (IaaS)
You wouldn’t refer to a donut shop as infrastructure. But the existence of that donut shop is possible because of the surrounding infrastructure: water, electricity, roads, bridges, railway tracks, and so on. With cloud computing, Infrastructure as a Service (IaaS) refers to the underlying components that allow software solutions to run. Examples of IaaS include the following components that can be provisioned to cloud customers:
Dedicated physical servers
Virtual machines
Virtual networks
Virtual firewall appliances
Virtual VPN configurations
Virtual firewall access control lists (ACLs)
Cloud storage
Platform as a Service (PaaS)
A platform refers to a conceptual workbench from which IT technicians can solve business problems with IT solutions—all without having to manage the underlying infrastructure.
You can deploy IaaS cloud resources individually, but when deploying more complex cloud services such as a web application, database, or big data analytics clustered solutions, with Platform as a Service (PaaS), the underlying infrastructure is configured for you; you only need to specify relevant details for what you are deploying. This is sometimes referred to as serverless computing; PaaS customers know that, in fact, there are underlying servers providing the solution, but they do not have to install, configure, or directly manage those servers. Another term for this is managed service. Examples of PaaS include
Microsoft Azure SQL Database
Amazon Web Services Relational Database Service (RDS)
Microsoft Azure Functions
Google App Engine
Other Cloud Service Models
Any type of cloud-based IT service packaged up into a manageable bundle is referred to as Anything as a Service (XaaS). Table 1-1 summarizes other common cloud service models.
TABLE 1-1 Other Common Cloud Service Models
CERTIFICATION OBJECTIVE 1.03
Identifying Cloud Deployment Models
Whereas cloud service models are categorized based on the service that each model offers, cloud deployment models are categorized based on where the cloud computing infrastructure physically exists, who owns and controls it, and who uses it. The following are the four common cloud deployment models:
Public cloud Cloud services available to anybody over the Internet, owned and managed by a CSP (any type of entity offering cloud computing services), such as Amazon Web Services, Microsoft Azure, Google Cloud, IBM Cloud, or a even a government agency offering online citizen services such as driver’s license renewals.
Private cloud Cloud computing infrastructure owned, managed, and used by a single organization, even if it is provisioned by a third party on-premises or off-premises.
Hybrid cloud The extension of an on-premises network into the cloud either during migration to the cloud or for the long term. Hybrid clouds can also combine private clouds (sensitive IT systems used only by one organization) and public cloud resources.
Community cloud Cloud services for a user base with the same computing needs, often driven by industry regulations. Examples would include government agency clouds that provide cloud-based software and security mechanisms required by government departments.
CERTIFICATION OBJECTIVE 1.04
Understanding Cloud Shared Responsibility
Cloud computing means using IT services running on infrastructure managed by somebody else. This is true even with a private cloud, which is usually controlled by a central IT department within an organization. Departmental use of private cloud services is tracked and charged back to individual organizational departments (departmental chargeback). But where should the line of responsibility for these IT services be drawn? The answer is, it depends on the specific cloud service, and even then, sometimes it is a shared responsibility.
Shared Responsibility Examples
The underlying networking, storage, and servers (IaaS) for a SaaS solution like web-based e-mail are under the control of the CSP, and as such, the responsibility for managing those items falls upon the CSP. Management means
Hardware acquisition and configuration
Firmware updates
Software installation, configuration, and updates
Maintaining service and data availability in alignment with SLAs
But what about the e-mail messages themselves? Contact information? Calendar entries? Who is responsible for the e-mail settings and archiving? Encryption? It boils down to this: cloud customers are responsible for managing any data they create. As always, there are exceptions, such as cloud providers encrypting data at rest in accordance with applicable SLAs.
Deploying an IaaS virtual machine means that the VM is under the control of the cloud customer, and as such, the management responsibility of the VM falls upon the cloud customer.
Deploying a managed database hides the infrastructure details from the cloud customer. The underlying infrastructure is the responsibility of the CSP, but the configuration of that solution and the management of the data stored in the database are responsibilities of the cloud customer.
EXERCISE 1-1
Sign up for a Free Microsoft Azure Cloud Subscription
In this exercise, you will sign up for a free trial of a Microsoft Azure account. To create a public cloud account, do the following:
1. Use your web browser to navigate to https://azure.microsoft.com.
2. Click the Free Account button, and then click the Start Free button twice.
3. Sign in with a Microsoft or GitHub account. If you don’t have one, you’ll need to create one to proceed.
4. On the About You page, fill in the required information, as depicted in Figure 1-5. Click Next.
FIGURE 1-5 Signing up for an Azure cloud account
5. On the Identify Verification by Card page, specify credit card information, which is used only to prove your identity; you won’t be charged. Then click Next.
6. On the Agreement page, check the box for the “I agree to the subscription agreement, offer details, and privacy statement” option, then click Sign Up.
7. After a moment you will be taken to the Microsoft Azure portal (the management GUI), as shown in Figure 1-6.
FIGURE 1-6 The Welcome to Microsoft Azure screen that appears after signing up
INSIDE THE EXAM
Hypervisors and Virtualization
Although not listed in the official exam objectives, the CompTIA Cloud Essentials+ CLO-002 exam expects you to be familiar with hypervisors and virtualization as they relate to cloud computing.
A hypervisor is a computer that can host multiple running virtual machines (this is one type of virtualization), also called guests. It does so by running virtualization software. The purpose of a hypervisor is to manage physical hardware resource access among run-ning guests.
Cloud computing depends on virtualiza-tion, but virtualization does not depend on cloud computing. Just because you are using virtualization, this does not mean you have a “gcloud.” Remember that cloud computing exists only when the six characteristics discussed at the beginning of this chapter (elasticity, scalability, etc.) are in place.
While a hypervisor itself can be run within a VM, this is done for testing purposes only and is not suitable for a production environ-ment. There are two common types of hyper-visors:
Type 1 hypervisors Also called bare metal hypervisors, this software is designed to run directly on physical server hardware. The hypervisor is the operating system, so an existing operating system is not needed. Type 1 hypervisors are used to run VMs in a production environment. Examples of type 1 hypervisors include Microsoft Hyper-V, VMware ESXi, and the open-source XEN hypervisor.
Type 2 hypervisors This virtualization software runs as an application within an existing operating system, so there is an additional layer between the hypervisor and the physical hardware. Type 2 hypervisors are commonly used by developers and IT technicians for testing purposes. Examples of type 2 hypervisors include VMware Workstation for Windows machines, VMware Fusion for macOS machines, and Oracle VM VirtualBox, which can run on Windows, Linux, Unix, and macOS (see Figure 1-7).
FIGURE 1-7 VMware Workstation, a type 2 hypervisor
CERTIFICATION SUMMARY
This chapter serves as an introduction to cloud computing. You have been introduced to the characteristics that define cloud computing: elasticity, self-service, scalability, broad network access, pay-as-you-go, and availability.
In addition, you have been exposed to cloud service models, such as Software as a Service (SaaS), Infrastructure as a Service (IaaS), and Platform as a Service (PaaS), that are used to provision resources via various cloud deployment models such as public, private, and hybrid clouds.
You also learned how the responsibility for managing different types of cloud resources sometimes falls upon the CSP and other times upon the cloud tenant. Finally, you learned about type 1 and type 2 hypervisors and how cloud computing relies on virtualization.
TWO-MINUTE DRILL
Identifying Cloud Computing Characteristics
Elasticity allows for the rapid provisioning and deprovisioning of cloud resources.
Scalability is the designing of cloud services to be able to increase or decrease virtual machine CPU and RAM amounts (vertical scaling) or to add or remove virtual machines (horizontal scaling) to support an IT workload.
Self-service means cloud users can provision and deprovision cloud resources using a GUI, command-line tools, or programmatically through API calls.
Broad network access allows a multitude of device types to access cloud services over a network.
Pay-as-you-go means cloud consumers pay only for the cloud resources they use.
Availability ensures that cloud-based IT systems and data are always available. The cloud SLA provides uptime guarantees.
Identifying Cloud Service Models
Software as a Service (SaaS) refers to software solution rentals where the software is accessed and used over a network.
Infrastructure as a Service (IaaS) refers to the underlying items such as storage, networking, and virtual machines that support cloud-based software solutions.
Platform as a Service (PaaS) refers to managed database and developer cloud services that do not require cloud users to configure the underlying supporting infrastructure.
Identifying Cloud Deployment Models
Public clouds are owned and managed by CSPs. The cloud services are available to anybody with Internet access.
Private clouds are owned, managed, and used by a single organization.
Hybrid clouds are composed of an on-premises network linked to a cloud virtual network.
Community clouds address organizations with the same computing needs, often within the same industry.
Understanding Cloud Shared Responsibility
With SaaS, the management of settings and data is the responsibility of the cloud customer. The CSP is responsible for managing the underlying infrastructure that supports the software solution.
With IaaS, the cloud customer is responsible for all management aspects of storage, network configuration, and virtual machines. The CSP is responsible for the underlying hardware, including routers, switches, storage arrays, and physical servers.
With PaaS, the cloud customer is responsible for the details regarding the PaaS solution, such as database settings and data. The CSP is responsible for the underlying infrastructure, such as storage and virtual machines that support the PaaS solution.
SELF TEST
The following questions will help you measure your understanding of the material presented in this chapter. As indicated, some questions may have more than one correct answer, so be sure to read all the answer choices carefully.
Identifying Cloud Computing Characteristics
1. Which cloud computing characteristic is most closely related to OPEX?
A. Broad network access
B. Elasticity
C. Pay-as-you-go
D. Self-service
2. Your manager has asked you to review the storage SLA for a public cloud provider to determine the potential amount of annual downtime. Which of the following cloud characteristics is most closely related to this scenario?
A. Broad network access
B. Pay-as-you-go
C. Scalability
D. Availability
3. Your cloud-hosted public website experiences more traffic during the holiday season. You need to design a configuration that responds to application requests to add or remove back-end virtual machines as required. The maximum number of VMs should never exceed four. Which cloud characteristic does this most closely relate to?
A. Metered use
B. Scalability
C. Broad network access
D. Availability
4. Users in your company use their work-issued laptops and personal smartphones to access web apps hosted on a company’s private cloud infrastructure. Which term best describes this scenario?
A. Broad network access
B. Scalability
C. Elasticity
D. Availability
Identifying Cloud Service Models
5. Which of the following cloud resource deployments is an example of IaaS?
A. Web e-mail
B. Instant messaging
C. Managed database
D. Virtual machines
6. You have decided to deploy your own cloud-based virtual machines hosting a Microsoft SQL Server database. Which type of cloud service model is this?
A. CaaS
B. PaaS
C. IaaS
D. SaaS
7. What type of cloud computing service model does cloud-based storage apply to?
A. SaaS
B. IaaS
C. PaaS
D. CaaS
Identifying Cloud Deployment Models
8. What makes private clouds different from public clouds? (Choose two.)
A. Service availability
B. Limited user base
C. Security of data at rest
D. Responsibility for infrastructure
9. Your on-premises network is linked to a cloud-based virtual network through a VPN tunnel. What type of cloud deployment model is this?
A. Hybrid
B. Private
C. Public
D. Extended
10. Which of the following is a true statement?
A. Anybody with Internet access can potentially access public cloud services.
B. Anybody with Internet access can potentially access private cloud services.
C. Private clouds are available to any user with a paid subscription.
D. Public clouds are used by a single organization.
Understanding Cloud Shared Responsibility
11. Which term best describes deploying a cloud-based database without having to configure the underlying virtual machine?
A. Horizontal scaling
B. Managed service
C. Vertical scaling
D. Infrastructure as a Service
12. You have manually deployed an Ubuntu Linux virtual machine in the public cloud. Who is responsible for applying Linux operating system updates to the VM?
A. Cloud service provider
B. Cloud tenant
C. Ubuntu
D. Cloud service provider and cloud tenant
13. Which type of hypervisor requires an existing operating system?
A. Type 1
B. Type 2
C. Type 3
D. Type 4
14. Which statements regarding cloud computing are correct? (Choose two.)
A. Virtualization relies on cloud computing.
B. Cloud-hosted virtual machines normally run on type 2 hypervisors.
C. Cloud computing relies on virtualization.
D. Cloud-hosted virtual machines normally run on type 1 hypervisors.
15. Which of the following is considered a “bare metal” type of hypervisor?
A. Type 1
B. Type 2
C. Type 3
D. Type 4
SELF TEST ANSWERS
Identifying Cloud Computing Characteristics
1. 
C. Operating expenses (OPEX) relate to paying for cloud resource usage, such as on a monthly basis, as opposed to hosting the same IT services on premises, which requires a capital investment in hardware, software, licensing, and technician fees.
A, B, and D are incorrect. They are not as closely related to OPEX as the pay-as-you-go cloud computing characteristic.
2. D. One of the details in a service level agreement (SLA) is the expected uptime (availability) for the cloud service.
A, B, and C are incorrect. They are cloud characteristics that do not relate to downtime.
3. B. Scalability refers to achieving elasticity through scalability configurations, such as the maximum number of virtual machines.
A, C, and D are incorrect. Metered use, also referred to as “pay as you go,” charges cloud customers based on their use of cloud resources. Broad network access does not involve adding or removing resources to improve resource performance. Availability ensures that cloud-based IT systems and data are available when needed and is normally addressed through backups and IT system and data redundancy.
4. A. Broad network access relates to the use of different types of devices to access cloud-based IT services over a network.
B, C, and D are incorrect. Scalability refers to the long-term design of constraints, such as the maximum number of virtual machines. Elasticity is the dynamic response to resource requirements. Availability ensures that cloud-based IT systems and data are available when needed and is normally addressed through backups and IT system and data redundancy.
Identifying Cloud Service Models
5. D. Virtual machines are considered Infrastructure as a Service (IaaS).
A, B, and C are incorrect. A and B are incorrect because they are considered Software as a Service (SaaS). C is incorrect because managed databases are considered Platform as a Service (PaaS).
6. C. If virtual machines are manually configured with software such as Microsoft SQL Server, this is considered Infrastructure as a Service (IaaS).
A, B, and D are incorrect. Communications as a Service (CaaS) relates to cloud-based services such as instant messaging. Platform as a Service (PaaS) provides services such as developer tools and databases without having to manually configure the underlying infrastructure. Software as a Service (SaaS) refers to a software solution available over a network, such as web e-mail provided over the Internet by a cloud service provider.
7. B. In the cloud, storage is considered Infrastructure as a Service (IaaS).
A, C, and D are incorrect. None of these cloud service models is considered IaaS.
Identifying Cloud Deployment Models
8. B and D. Private clouds are the responsibility of and used by a single organization, but still adhere to cloud computing characteristics such as self-service and elasticity.
A and C are incorrect. Service availability and security of data at rest are items that apply to private and public clouds.
9. A. Hybrid cloud solutions combine on-premises and cloud solutions, such as linking an on-premises network to a cloud network via a VPN tunnel.
B, C, and D are incorrect. Private clouds are the responsibility of and used by a single organization, whereas public clouds are potentially available to anybody over the Internet. “Extended” is not a valid cloud deployment model.
10. A. Private clouds are the responsibility of and used by a single organization, whereas public clouds are potentially available to anybody over the Internet.
B, C, and D are incorrect. Private clouds are owned and used by a single organization. Public clouds are available to all Internet users.
Understanding Cloud Shared Responsibility
11. B. In the cloud, a managed service provides a solution, such as deploying a database, without having to deal with the underlying virtual machine, network, and storage configuration details.
A, C, and D are incorrect. Horizontal scaling adds or removes virtual machines to support an application. Vertical scaling increases or decreases virtual machine power through items such as the number of virtual CPUs and the amount of RAM. Infrastructure as a Service (IaaS) requires the detailed configuration of resources such as virtual machines, storage, and networking.
12. B. Cloud tenants are cloud customers, and as such are responsible for applying updates to manually deployed virtual machines.
A, C, and D are incorrect. These entities are not responsible for applying updates to a manually deployed virtual machine.
13. B. Type 2 hypervisors require an existing operating system since they run as an app in the operating system.
A, C, and D are incorrect. A type 1 hypervisor installs directly on hardware; it is the operating system. Type 3 and type 4 are invalid hypervisor categories.
14. C and D. Virtualization in the form of virtual machines makes cloud computing possible.
A and B are incorrect. Virtualization can be used outside of a cloud environment. Cloud-hosted virtual machines run on type 1 hypervisors (bare metal) in the cloud.
15. A. Type 1 hypervisors run directly on hardware (bare metal) and do not need an existing operating system. They are often referred to as a bare metal hypervisor.
B, C, and D are incorrect. Type 2 hypervisors run as an app and require an existing operating system. Type 3 and type 4 hypervisors do not exist.