What are Data Centres, and Why Do Companies Need Them?

Nothing New Under the Sun, or a Short History

ENIAC, the first or first “generation” computer, was introduced to the world on February 16, 1946, having undergone construction a year prior. The first large computers then began to appear in the earlier 1950s. Large, expensive, and noisy, these tube (and later, transistor) monsters generated massive amounts of heat and had to be operated in large, specially cooled halls. From today’s point of view, they used gigantic magneto tape, drum, and disk memories to work with data. For example, the IBM 2311 disk drive from the 1960s was about 1x1 meters in size, used removable disks 610mm in diameter, and had a capacity of only 7.25 MB.

In the 1960s, computer rooms began to be gradually standardised. Power distribution was routed through the floor, air conditioning systems were developed, and access security was improved. As technology improved, minicomputers and eventually personal computers were developed, local computer networks were invented, and companies began building their own server rooms, as well as rooms for connecting servers to local networks. Most of them were so-called on-premises solutions, meaning they were located on company property. The explosion of the Internet and the “dot-com” boom of the 1990s then sparked major demand for web servers and hosting services. And so the first commercial data centres began to emerge to give companies a better place to run their servers. These data centres offered higher capacity, connection to high-speed fibre networks, and most importantly, peace of mind amidst the increasing demand on IT system operation.

The End of the Server Room Era

However, corporate IT infrastructure is no longer just about server rooms and cables distributed around the building. Reliable and secure data backups, protection against cyberattacks, uninterrupted energy supply, device availability, and high-speed Internet connectivity must all be ensured when operating corporate IT. All this requires a sizeable initial investment as well as maintenance and operating costs. And in today’s chaotic and unpredictable cyberspace, constant investment in updates, modernization, and staff training is a must. Any infrastructural expansion, no matter how small, often means new, considerable expenses. For this reason, data centre services can be a viable and economically advantageous solution.

Are Data Centres Really Beneficial?

It is often said that by placing servers and technology in data centres, we place them in a safe environment without losing control. However, as we know, nothing can be perfect, and using data centres certainly has its pros and cons. By using data centre services, a company can gain:

• prostředí navržené a optimalizované pro provoz serverů a dalších IT technologií,
• spolehlivé napájení (více napájecích elektrických tras, UPS, dieselové generátory),
• odbornou správu a nepřetržitý dohled nad technologiemi, a tudíž méně starostí,
• spolehlivé a bezpečné vysokorychlostní připojení do internetu,
• kromě fyzické (přístupové systémy, kamery, dohled) i vysokou kybernetickou bezpečnost (např. antiviry, firewally, obranu proti DDoS útokům a další),
• bezpečné a obvykle i redundantní zálohování dat (geograficky oddělená datová úložiště),
• vysokou dostupnost služeb (např. založenou na SLA),
• jednodušší škálovatelnost (při změnách není třeba investovat do rozšíření provozního prostředí),
• eliminace rizik z nepředvídaných událostí (požáry, povodně, živelní pohromy),
• možnost napojení infrastruktury na cloudové služby a hybridní modely.

While moving IT technology to a data centre does not mean losing direct control, device access can become somewhat more complicated. It may be necessary to drive to the physical data centre, pass various security checks at the entrance, and follow specific regulations once inside. On the other hand, many data centres provide, for example, facilities for setting up and configuring equipment, storing small materials on-site, or resetting or shutting down equipment directly through the operator.

The quality of connectivity between the company and data centre can also play an important role, especially when running time-sensitive applications.

Financial Aspects

Of course, transitioning to a data centre will change the process of financing corporate IT. Energy, air conditioning, and service costs will decrease, while regular monthly payments for the data centre will need to be integrated. IT cost planning will be simplified, and most of the costly, long-term upgrades to server room equipment will be eliminated. Whether the use of data centre services will be financially advantageous in the long term and to what extent must thus be evaluated by company management.

However, an evaluation of this manner must include not only financial aspects but also issues of flexibility and future development of the company. It is much easier to add a few servers to a data centre than to expand a company server room, not to mention the fact that IT staff will finally be able to devote themselves to development instead of never-ending IT maintenance.

Small- and Medium-sized Companies

Moving IT to data centres has already freed many small- and medium-sized companies from everyday worries and emotional disputes concerning IT management. In particular:

• An environment designed and optimized for server and other IT operations,
• Reliable power supply (multiple supply lines, UPS, diesel generators),
• Expert management and continuous supervision of the technology, and therefore fewer worries,
• Reliable and secure high-speed Internet connection,
• Both physical (access systems, cameras, surveillance) and cyber security (antivirus protection, firewalls, DDoS defence, and more),
• secure, event redundant, data backup (geographically separated storage),
• high availability of services according to SLAs,
• easier scalability, without the need to invest in expansion amidst changes,
• eliminated risk of natural disasters and unforeseen events (fires, floods, accidents),
• the ability to connect infrastructure to cloud services and hybrid models.

Tiers and More

When evaluating the level of reliability and availability of data centres, we often come across the notorious “Tier classification” system. These Tiers represent an internationally recognized standard, created by the Uptime Institute, for evaluating the physical infrastructure of data centres. The level of reliability and availability is divided into four categories according to the number of electrical connections, system redundancy, availability, maximum permissible annual outage length, maintenance options, and fault tolerance. Very briefly and simply (plenty of detailed descriptions and tables are available online), we can summarise the Tier classifications as such:

• Tier I: centres without redundancy, operation with planned outages
• Tier II: partial redundancy, higher reliability than Tier I but still with outages
• Tier III: full redundancy (N+1), operational maintenance without outages, high availability
• Tier IV: highest resilience and availability, full redundancy (2N), resistance even amidst equipment failures

However, this tier system is not the only way to evaluate the properties of data centres. Other verified and utilised certification models include, for example, the European standard for data centres (EN 50600) or the ISO 22301 standards for business continuity management and process resilience. These are summarized in more detail in the table below.

Of course, there are several other important accompanying certifications that data centre service providers should have or meet. These include norms and standards in the field of quality management, environmental management, and service management in the fields of information technology, energy management, and last but not least, infrastructural compliance with legislative requirements, such as GDPR, NIS 2, or others. It is thus advised to ask a potential provider about these certifications before concluding a contract.

Choosing a Data Centre

Numbers and certifications are certainly important guides when choosing a suitable data centre, but other “minor” details also play a role, such as whether the provider operates from its own data centre or only leases part of a larger data centre to provide services indirectly. In addition, it is important to consider how data backups are organised and whether backups can be geographically separated from each other, how safety and security is ensured, what level of cybersecurity is provided, and whether it is possible to easily link the infrastructure to other IT services, such as data network services, IoT networks, and telecommunication services, as well as to cloud and other services not exclusively IT-related.

Customer service represents another basic, though often neglected, criterion when choosing data centre services. Although anything can be said on the Internet, reality is often less ideal, and the consequences of hasty decision-making can be unpleasant, to say the least.