A Little Guide to the Big World of Cloud

From Punch Tape to Cloud-Native IT

If we take a closer look at the relatively short but rapid history of computing, we will find that it consists of a sequence of repeated decentralisations and recentralisations of computing equipment, or, as they are often called today, resources. The first commercial computers were given tasks via paper punch tapes and, later, punch cards, which users carried to computing centres to be processed. The results were then printed, usually the next day, on perforated paper and known as tape listings.

These annoying visits to computer centres were soon replaced by character terminals, devices with keyboards and monitors connected by cables to mainframe computers. This made it possible to feed a computer from a remote workplace while also displaying calculations. Some computers in networks, known as servers, were later set aside for data storage or running specific applications. Gradually, computer networks with centralised resources were then created. Lastly, with the growing amount of data and applications in individual computers and the pressure for users to share and collaborate, computer and data networks were also developed.

With the advent of personal computers in the 1980s, history more or less repeated itself. The idea that not everyone would have a personal computer in their workplace was quickly replaced by new coworking and data sharing standards. The first local networks with equivalent computers, so-called peer-to-peer networks, were created. Users of these networks would share data amongst each other, typically at their own discretion. However, as they grew in scale, peer-to-peer networks became more and more complicated to manage. Complex, data-intensive applications were logically moving to increasingly powerful servers, and personal computers were evolving faster and faster to the level of intelligent terminals. Client-server architecture and concepts such as low-cost thin clients began to dust off old network principles from the 1970s. From this, a whole spectrum of new architectures was created, many of which, such as web interfaces, are still used today.

Centralization in local networks had two major consequences. First, communication between computers and servers grew rapidly, which, among other things, fueled the increasingly rapid development of network technologies. Second, capacity and performance requirements for both networks and the devices that serve them grew significantly. Server rooms were filling up, and the amount of energy and service needed to power them was increasingly sharply. The constant updates, upgrades, and cybersecurity requirements had become too economically and mentally taxing for many companies, who were now looking to get rid of all IT management.

Meanwhile, along with the commercialization of the Internet and the widespread availability of high-speed Internet connection, another so-called decentralised centralisation occurred. In the 1990s, the first commercial data centres were created, allowing companies fed up with IT operation anxieties to put away their server room equipment and leave others to take care of it. Data centre operators then begin to ask themselves, quite logically, why they should only rent space in data centres when they could instead rent entire servers and other IT infrastructural elements. This would allow them to help clients without having to invest in their own hardware or software. But there’s a catch. The average utility of typical corporate servers rarely reaches 30%, though there are moments, such as during the financial closing period, when the server load can reach up to 100%. Renting out such seldom-utilized servers probably would not be worth it. At this point, virtualisation and the word cloud come into full play. The year is now 1997.

Virtualisation is nothing new under the sun, either. It was used in mainframe computers, which could perform dozens or even hundreds of tasks at the same time. Virtualisation is also commonly used by our well-known Windows, as each open window is actually nothing more than a separately running task or application. In practice, virtualisation means that the cloud provider does not lease the services of actual servers but virtual servers, which only operate as separate applications. In simple terms, several virtual servers can run simultaneously on one physical server. Optimal utilisation and load balancing between individual physical servers is then ensured by virtualisation platforms.

Cloud, Complications, and Caveats

Applications are becoming more and more sophisticated, and most of us likely cannot even imagine the petabytes, exabytes, or zettabytes of data that will eventually be stored in the cloud. The industry bottleneck will thus no longer be the performance of cloud servers but the ability to access to their services.

In the past, problems accessing services have been solved mainly via busy web servers. Frequently used data, e.g. frequently visited websites, is therefore not provided directly from web servers but is located closer to the user on so-called proxy servers. From there, data are delivered to users. Proxy servers then turn to their own service only if they do not have the required data available. Multiple proxies can serve a single cloud service, significantly reducing their load. A similar principle is used for media streaming or TV distribution.

IoT and Cloud

Data is another big problem. Billions of different sensors and detectors churn out gigantic amounts of raw data, which need to be stored and often processed in real time, especially through IoT networks. Similarly, another decentralisation is occurring, this time referred to as edge computing. To some extent, edge computing architecture can be compared to proxy servers, as it moves data processing as close as possible to its sources without first cramming and processing it in a cloud data centre. Instead, processing takes place at the edge, as close as possible to the source or data consumer. Only the processed results, typically much smaller in volume, are then transferred to and stored in the cloud. The advantage of this approach is mainly faster processing, lower latency, and a smaller load on transmission networks, which also translates into lower operating costs.

IT technologies are rapidly evolving and changing, both in general and within the cloud itself. New trends are coming, such as containerisation, Cloud-Native IT, or infrastructure as code[1], which many see as the future of information technology.

Cloud Computing is When…

According to the U.S. National Institute of Standards (NIST)[2], cloud computing is a model for IT service providers that allows convenient network access to a group of shared computing resources (e.g. networks, servers, data storage, applications), which can be provided or freed up quickly with minimal administrative effort or cooperation with the provider. It is characterized by five basic features:

• On-Demand Self Service: A client can order services on their own without needing to interact with the operator.
• Broad Network Access: The network uses standard mechanisms and platforms.
• Resource Pooling: Resources are shared by dynamic allocation and redistribution among multiple users.
• Rapid Elasticity: Resources can be used to scale quickly, both expanding and releasing.
• Measured Service: Both service providers and customers may manage, optimize, and take count of resources used.

It is worth noting that although the above definition dates back to 2011, it is still valid, has not been replaced by any newer version, and is still widely accepted as a frame of reference.

In short, the basic feature of the cloud is the fact that IT services are separately operated by the provider off the client’s premises and are provided as a virtual service.

IT as a Service

Through the cloud, a plethora of virtual IT services are possible. With some exceptions, such as real-time activities, the cloud can do anything that on-premises IT services can, and often much more.

Cloud services can basically be divided into three main groups:

• Software as a Service (SaaS), a model where software applications are run as a service in the cloud. Users typically connect to the service through a web interface, a client on a computer, or an application programming interface (API). Typical examples of this are office applications such as Microsoft 365, Google Workspace, or accounting software.
• Platform as a Service (PaaS), providing clients with the ability to run their own or adopted applications on a virtual cloud platform. The platform (hardware, operating or runtime systems, and basic software) is managed by the provider. PaaS is often used by developers to develop, test, deploy, and manage applications.
• Infrastructure as a Service (IaaS), a cloud model that provides clients with virtual computing resources like servers, memory, networking environment, and other necessary components to create their own virtual IT infrastructure. IaaS relieves the client of the need to invest in their own hardware and manage their own server rooms or data centres.

As we can see, the cloud is nothing more than an effective combination of virtualization, outsourcing, and the use of modern communication technologies, which make it possible to effectively shift IT worries onto the shoulders of specialised organisations.

Advantages and Disadvantages of Cloud Solutions

Just like with everything, the use of cloud services has its advantages as well as its disadvantages. Many providers argue a mantra that the cloud is cheaper because clients do not have to purchase hardware and oversee the operating environment for their IT infrastructure. This may often be the case, but it is far from the rule. First of all, the transition to cloud cannot be done with the snap of a finger. In reality, it is a very serious and important decision that must fit into the client’s overall corporate strategy.

In the case of platform and infrastructure cloud services, for example, it is important to consider that the possibility of influencing cloud infrastructure will always be somewhat limited. While clients can literally do whatever they want with their “home” infrastructure, with the cloud, they are always limited by the rules outlined by the provider. The demands of applications on the cloud and response time, the connection between cloud services and internal IT that cannot be migrated to cloud, and many other, seemingly miniscule details must be evaluated beforehand.

It is extremely important to choose the correct provider, i.e. one who does not make false promises but discusses all aspects of cooperation with potential clients in a well-founded and responsible manner, prepares a transition project, and participates in its implementation. While a move to the cloud is not irreversible, returning to on-premises IT or, in the worst cases, moving to another provider may not be entirely easy.

Is the Cloud Secure?

While this discussion has been ongoing for years, it remains puzzling in nature. The cloud is only as secure as its provider makes it and only as secure as the connection through which a client access cloud services. The important thing is that few companies or organisations can afford to invest in security, either physical or cyber, but reliable cloud service providers will already have these at their disposal—and at a fraction of the cost for individual clients.

Slowly but Surely

It doesn’t always make sense to start a grandiose project and move the entire company infrastructure to the cloud right away. You can also start slowly, moving data backup functions or company data storage to the cloud. These tasks represent much smaller projects, which can give an organisation experience working with the cloud and cooperating with the provider. If during these projects, the provider does not meet the client’s expectations, the provider can easily return the services back to the company or move them elsewhere.

Cloud a Startups

For startups or fast-growing smaller business looking to pinch pennies, cloud infrastructure can relieve significant financial and operational pressure. These clients enjoy not having to invest unnecessarily in relatively expensive hardware and software licenses, instead purchasing cloud infrastructure that meets the requirements for administration, data processing, and security for a reasonable monthly fee. Most importantly, they do not have to worry about managing their own IT services and can fully devote themselves to running their businesses.

What You Can Find in the Cloud

Few people can imagine all the different services that can be found in the cloud. Most people’s ideas about the cloud end with providing data storage for family photos from mobile phones, office software, or tools for backing up company data.

České Radiokomunikace offers modern and secure solutions beyond cloud-based corporate IT infrastructure. Let’s take a closer look.

CRA Business Cloud

This is a Czech cloud for powerful, reliable, highly accessible, and secure cloud IT infrastructure operated in an environment equipped with both traditional cloud services and the latest solutions, such as Kubernetes or Cloud-Native IT. By “Czech,” we mean that all data and applications are located in the Czech Republic, home to strict legislative requirements. With CRA Business Cloud, clients fully manage their virtual infrastructure through our self-service VIRTIX customer portal.

Data Backup and Restoration

Backing up and restoring data or all IT functions to cloud storage ensures that clients’ businesses do not stop running, even in the event of catastrophe.

Data storage and archiving

Storing or archiving large volumes of data not only guarantees clients that that they will never lose their information but that they will also have quick, easy access to it at any time.

Cloud and the Internet of Things (IoT)

Through CRA’s LoRaWAN IoT infrastructure, clients can connect their devices, sensors, meters, and other elements directly from the field to an application or digital system. However, CRA goes even further in this respect, providing a universal edge computing platform unique to the Czech Republic and with enormous synergy potential known as MultiTech IoT Cloud.

Streaming Services and Media Cloud

CRA Media Cloud provides an economically advantageous solution for streaming and OTT services, capable of processing, storing, and delivering video content to all devices from mobile to TV and built within a unique cloud infrastructure.

AI as a Service

Even artificial intelligence can be provided as a cloud service. CRA AI Cloud will provide both space for the operation of its own AI solutions as well as a number of pre-trained AI models prepared for a variety of applications.

[1] https://www.cra.cz/budoucnost-cloudu-je-kontejnerizace-a-it-infrastruktura-jako-kod
[2] https://csrc.nist.gov/pubs/sp/800/145/final