A reliable data center is the foundation of the cloud

Cloud services for their full-fledged operation require a special server infrastructure, which is easiest to provide with the help of so-called data centers (synonym – data processing center). A data center is a generic name for specially equipped premises where data servers, network equipment, and other components for their operation are located. The main task of a data center is to ensure an uninterrupted operation of servers where data is stored and processed as well as network infrastructure. The advantages of hosting in data centers are clear – it is easier to maintain the performance, fewer infrastructure problems, load separation, etc. Therefore, today almost all companies that need to process large amounts of information use data centers for work.

A data center infrastructure and components

A modern data center is a combination of such elements:

• information infrastructure – server equipment, data storage, etc. The role of the information infrastructure – data accumulation and processing;
• communication infrastructure – network equipment and communication channels. Thanks to network infrastructure, data is sent to the data center, and distributed between its nodes, and transmitted to users;
• engineering infrastructure – systems that ensure an uninterrupted existence of a data center. This includes backup power supply systems, air conditioning and maintaining stable temperatures and humidity, fire alarms and emergency fire extinguishing systems, access control, etc.

 

Depending on the purpose of a data center, its size, the degree of data security, its structure may include other systems, for example, data theft protection, video surveillance, seismic events alarm system, etc.

The main criteria of a data center operation quality

A data center operation is usually evaluated according to the following criteria: uptime, reliability, fault tolerance.

 

Uptime is the percentage of system availability for a certain period of operation (for example, per day). Uptime depends on many conditions – a communication channel size and its workload, equipment failures, power interruptions, overheating of system components, etc. For serious data centers, the uptime indicator is higher than 97%, and in some cases, it is higher than 99%.

 

Reliability and fault tolerance is the ability of a data center to perform its functions, even if the operation of one or more components is disrupted or other accidents have occurred. For example, the power supply has been cut. Data centers strive to increase fault tolerance by duplicating each element – communication channels, data storage devices, uninterruptible power supplies, spare servers with automatic switching in case of problems.

 

Since data security and server availability are the main criteria by which the client chooses a data center, serious data centers pay much more attention to these aspects. Most often, the American standard TIA-942 and the rating scale, which classifies data centers by reliability, is used for comparison:

• Tier 1 – there is no backup systems that ensure operability, a failure in one of the nodes and subsequent repairs lead to a complete shut-down;
• Tier 2 – there is a backup power supply and other backup systems for ensuring the data center operation, but its repair requires stopping the equipment operation;
• Tier 3 – the central network, engineering, and information elements have reserves, repair work is possible without disconnecting equipment, all systems are reserved once;
• Tier 4 – complete duplication of all systems, reserves are provided for both main and spare elements – several cooling systems, 2-3 UPS, etc.

"Trash data center" — economy or threat?

In recent years, the concept of "trash data center" has appeared. This name defines data centers where there is no even a minimum infrastructure, which is required for Tier 1. Usually, this is a non-specialized building (warehouse or hangar), where racks with servers are installed. Most frequently, customers receive the most basic package of services, where even a constant power supply is not guaranteed, you can optionally install your UPS. The package also includes an Internet channel and a server.

 

The main advantage of trash data centers is low cost, placement of a server here is several times cheaper than in high-quality data centers. At the same time, the client is not guaranteed high uptime or even data security – everything is at their own risk. Perhaps for someone, such a cooperation model will be acceptable, but for serious customers, this is a great danger, and such data centers should be avoided. The easiest way to distinguish unscrupulous providers from normal offers is to pay attention to some aspects:

• extremely low price;
• complaints on the web about equipment operation;
• low uptime or lack of information about the acceptable time of services unavailability;
• absence of any guarantees from the data center.

 

Placing a company's infrastructure in the office or utility space can be also considered as an example of a trash data center. All the aforementioned examples, with few exceptions, have all the pre-mentioned risks.

The data center of the cloud service provider Tucha

To provide customers with the best services and with guaranteed access 24 hours a day, the capacities of the cloud provider Tucha are located in the large data center Telehouse Frankfurt (Frankfurt am Main, Germany). This location was not chosen by chance – we took into account the proximity of one of the largest traffic exchange points in Europe, DE-CIX, which ensures fast and uninterrupted data transfer between data center elements, as well as between equipment and customers. Also, the data center is directly connected to the Frankfurt fiber-optic network.

 

The data center covers an area of 67 thousand square meters, 25 thousand of which are allocated for hosting servers. The power supply is provided by several lines of 8 kW each, in addition, a mandatory backup power supply is provided. Climate control is provided by an advanced air conditioning system and sensors that ensure the maintenance of a temperature at the level of 24 degrees and humidity at the level of 15-50% all year round. All this allows you to provide the highest level of data security, high uptime, and fault tolerance. The reliability level of the data center according to the American classification is Tier 3.

Conclusions

The operation of online services depends heavily on the choice of a provider. Therefore, if you are going to conclude a contract with a particular company, be sure to ask which data center services it uses. After all, the speed, reliability, and fault tolerance of the selected services, and consequently your applications' availability, largely depends on its infrastructure.

 

We have chosen the best data center to provide the highest quality services. To see this, visit the page with the reviews from our customers and contact us for competent consultation. We are always in touch!