An introduction to Infrastructure as a Service (IaaS) without the hype. What is IaaS? How do businesses use IaaS today? Let’s keep it simple.
“Cloud” is just a server made available via the Internet
Cloud computing is more confusing as a term than as a real-life object. Let’s cut through all the marketing jargon to understand what AWS, Microsoft or Google means when they say “in the cloud.”
“Cloud” is just a server made available via the Internet to perform services. When that server sits in your office under your responsibility and direct control, it’s called a private cloud. When it’s located in a data center, it’s a public cloud.
IaaS (and all traditional as-a-Service solutions) refer to public cloud lease agreements.
Understanding Infrastructure as a Service and Other Cloud Products in traditional business terms
Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service share some remarkable similarities with traditional industries. Understanding these similarities can help anyone start to make sense of their differences.
In every industry, businesses position at various segments of the value chain. From manufacturer to retailer, goods undergo a process of refinement and customization until they can be consumed. Price increases with each process step, downstream as each segment adds value, making it easier for their customers to use the product.
Computing is quite similar. The semiconductors industry manufactures the base product and sells hardware to cloud providers. The provider builds its hardware into a data center and leases out use of its hardware.
Cloud providers are like wholesalers. Like any good wholesaler, they sell to a variety of customers. Some of their customers are actually retailers who then turn around and sell to the end-user. They might also, in the case of Microsoft and Google, develop software that they sell directly to consumers.
Where does your business fit? You are the retailer. But unlike a traditional retailer, many of the services you provide will be used by your employees, and you might not charge your customers to use your services.
What does as a Service mean?
“As-a-Service” denotes the lease agreement between cloud providers and their customers.
What is the difference between IaaS, PaaS and SaaS, BMaaS, and serverless?
As-a-Service (aaS) offerings differentiate on how much software and management is included in the lease agreement.
Bare Metal as a Service provides direct access to underlying hardware to allow customers the flexibility (and responsibility) for determining their own virtualization needs. In terms of management and upkeep, BMaaS is the closest approximation to private cloud.
What is virtualization?
Virtualization is allocation of compute resources. When using a PC your Operating System makes these kinds of decisions for you easily because typically one process occurs in the foreground while others hold in the background. But when business functionality is on the line and multiple processors, hard disks, and networks in play, you need to set rules and boundaries on the people, processes, and tools that receive computing resources.
Infrastructure as a Service agreements lease virtual machines that come with allotted compute, storage, network bandwidth, and Operating System installed. Once bought, developers will install the languages, frameworks, and packages they will need, then develop and deploy services on the platform they set up. IaaS accounts for the second-most revenue of all the cloud services, behind Software as a Service, and growth for this service continues to be strong.
Platform-as-a-Service offerings specify the end use of the software more so than IaaS. Cloud providers like AWS and Azure typically include many different types of PaaS ready for lease and already geared for specific projects. Common PaaS solutions include Machine Learning, Web Applications / Websites, Business Analytics, Databases, and more.
Each of these end-uses has specific computing requirements that are built into the PaaS solution. For example, many Machine Learning (ML) frameworks work best with Graphics Processing Units (GPUs) not CPUs. Thus, PaaS ML solutions frequently offer higher GPU compute than others.
Languages, frameworks, and developer tools will often be pre-installed onto the PaaS, so that developers can spend even less time on setup and management. The downside of this is lock-in. Your teams will require some ramp-up time to learn languages and frameworks if you want to switch providers, and the software you develop will require the software included in the PaaS.
Serverless is an emerging cloud service that allows developers to deploy containerized applications on an as-needed basis. Containerized applications come with everything they need to run. Containers plug and play with any Operating System, so cloud providers have more flexibility to provision and run their customers’ containers where and how they want to achieve the most efficiency. They pass some of these efficiencies on to their customers by charging a true price-per-use.
What is an example of Infrastructure as a Service?
Every cloud provider we know of sells IaaS, but here are some of the most common.
What are some use cases for IaaS?
- Disaster Recovery (DR) – Virtualized DR benefits from scalability, consolidation of on-premises resources, and diversification.
- Testing and development – Developers frequently use IaaS to build and deploy Software-as-a-Service and web applications
- Big data – Storing, managing, and analyzing both structured and unstructured data requires scalable compute power, disk space and network bandwidth. Aggregating disparate data sources is also best done in a centralized location (Ie. the cloud).
- Add-on services – IaaS enables businesses to install an array of solutions on top of their business computing infrastructure. Security, business analytics and intelligence are common.
Is IaaS secure?
The idea of migrating private on-premise data and applications to the cloud still causes concern, but without data-driven reason. Most security breaches in the public cloud have been attributed to misconfiguration, and fewer breaches have occurred than in on-premise installations (Gartner, 2020). In other words, most security breaches have been caused by poor security practices on the part of the customer, not the cloud provider.
Security in a public cloud environment is shared responsibility. The customer has the freedom and responsibility to determine how users will connect to their services. Remember that all cloud services are available via the Internet. IaaS does need to be configured properly with the right services and protocol to maintain security.
In other words, Public cloud offerings are like building your castle in front of an impassable mountain. You can be reasonably sure that your back is covered. But you absolutely need to maintain the defense of the front and sides.
Considering the steep increase in cyberattacks over the past two years remember these two rules for cloud security:
- When choosing a cloud provider, always consider security.
- When deploying and maintaining IaaS, always consider security.
Closing Remarks
We hope you have learned a bit more about cloud computing, and most importantly that you feel more comfortable with the variety of cloud services and how they might suit your needs. If you have any further questions please email david@platteriver.com for more information.
Additional Reading
Panetta, K. (2019, October). Is the Cloud Secure? Smarter With Gartner. https://www.gartner.com/smarterwithgartner/is-the-cloud-secure/.
Worldwide Public Cloud Services Market Totaled $312 Billion in 2020 with Amazon Web Services and Microsoft Vying for the Top Position Overall. IDC. (2021, May). https://www.idc.com/getdoc.jsp?containerId=prUS47685521.
