What is Cloud Computing

Overview

Teaching: 10 min
Exercises: 0 min

Questions

• What is cloud computing and how can I use it?

• What are common cloud services available?

Objectives

• Learn some historic developments that lead to cloud computing

• Understand the main drivers of cloud computing

• Know typical use cases of cloud computing relevant for your work

In the early days of computing, computers were called mainframes and filled a big room. To let the perform computations, you have to feed them punch-cards, so most of the time it’s users were busy sitting behind a desk punching holes in pieces of paper. When they were done, they would leave the punch cards in a mailbox, and after some time, receive the results of the computation (if no error occurred) printed on paper from another mailbox. Since these computers were very large and expensive, many different users

For the users of these computers, the computer terminal with a screen and keyboard was a very welcome innovation. As computer were still expensive, multiple terminals were connected to the same mainframe. All data was stored and computations happened on the mainframe, with the terminals only serving as endpoints providing a keyboard instead of punch cards and a screen instead of text printed on paper. In modern day computing, we still see the word terminal appear, but usually as an application that can be used to interact with a computer system.

By the 1980’s computers got small enough to sit on a desk, and personal computers emerged, machines not shared by many users but dedicated to a single person or family. For some decades, this was a common way people used computers. Even though big companies still used centralized main-frames for important databases, for most people networking was still expensive and most computers were used in a stand alone way and data was most commonly transferred between computers using physical data carriers.

In the first decade of the 2000’s, increasing access to (wireless) internet and the introduction of smart phones resulted in an explosion of more centralized services, such as web-mail, maps and photo albums. This made it easier for users to access their data and the same service from different computing devices, without the need to transfer data between devices using physical storage. An added advantage of this centralization is that you have a single truth. If you have two different versions with on two different devices, it can be difficult to know which one is the best one, but with a centralized approach you have a single truth.

The advantage of this kind of centralization is that it becomes possible to outsource the configuration and management of computers and software. While it is possible to set up your own private mail service, doing so takes a lot of effort, so most individuals outsource this to Google, Microsoft or their Internet Service Provider. The ability to outsource complicated IT management tasks to companies who have a lot more knowledge and experience is potentially very attractive to many (smaller) companies, as they can avoid having to set up their own IT department and focus on their core business competencies.

While access to the internet became more and more widespread, computers became so powerful that their resources were mostly idle during daily office work, and operating systems were improved to be able to run many different programs in parallel. This innovation was taken to the extreme by the development of virtualization software which made it possible to let a big and powerful computer behave as if it were a number of smaller virtual computers, allocating computational resources to the virtual computer that requires it the most. This provided significant economy of scale benefits, in particular to companies that already had big data-centers full with computers providing services to internet users.

Arguably, the advantages offered by centralization, the easy of use thanks to outsourcing and the scalability offered by virtualization, are all important drivers of the success of the cloud. In the last decade, a plethora of cloud services offered by all kinds of businesses has emerged. As of 2021, major cloud service providers such as Amazon Web Services and Microsoft Azure offer hundreds of different cloud services to their customers. To make sense of this scala of services, they are often categorized by Something-as-a-Service labels. Full featured, ready to use services like Google Docs, Office 365 or Overleaf are often denoted by the Software-as-a-Service label, abbreviated to SaaS.

In this workshop, we focus on the other end of the spectrum, often denoted as Infrastructure-as-a-Service (IaaS). An important part of this is the ability to rent a (virtualized) computer running in the cloud with little effort, paying only for the time the computer is running. It is like renting a server that you can use for whatever purpose you see fit, paying only for it while it is turned on. This particular type of service is typically called a Virtual Machine (VM) or Virtual Private Server (VPS), and is offered by most cloud providers. Many of the other cloud services are built on top of this type of service, and typically focus on a more specific task such as Data Science, Machine Learning, File Storage, Databases, Web Services, etcetera. Many of these types of services consist of a VPS with some additional software pre-configured to facilitate a particular task, i.e. a Data Science cloud service cloud consist of a VPS that comes with R and Python and the most popular packages pre-installed, saving you the hassle of having to install them yourself.

Some cloud providers

Below is a list of some well-known cloud providers and useful services offered by them.

• Microsoft Azure offers Virtual Machines and has a free $100 trial for students
• Amazon Web Services offers virtual machines under the name Elastic Compute Cloud, abbreviated to EC2. They have a free tier that provides the option to run virtual machines with few resources for up to 750 hours.
• Google Cloud offers virtual machines under the name Compute Engine and provides $300 free credits to new customers.
• IBM Cloud offers server under the name IBM Cloud Infrastructure. They may offer discounts or free services through their Academic Initiative Program.
• Alibaba Cloud is a major cloud provider in China

There are many other smaller providers that provide VPS or IaaS kind of services. Furthermore, other types of cloud services that can be interesting to consider are:

• Google Colab is a specialized cloud service that focuses on writing and running Jupyter notebooks for free, including access to GPU’s.
• RStudio Cloud is a specialized service for running RStudio and R on a cloud instance, which can be used from your browser.

Key Points

• Explain concepts of centralization, outsourcing and virtualization

• Discuss different cloud service models such as IaaS and SaaS

• Name the important cloud service providers such as Amazon and Microsoft