In this article you will read about,

• how many data centers T-Systems operates,
• which measures ensure resilience and sustainability,
• and which trends will be important for the data center of the future.

Around 10 years ago, T-Systems laid the foundations for one of Europe's most powerful data centers in Biere. About a year and a half later, the site began operations. And three years later, the Open Telekom Cloud found its home there. Data center expert Johannes Krafczyk has been involved in the construction from the start. To mark the anniversary, the T-Systems representative at Bitkom has compiled exciting facts and figures about the Biere site, green IT and the data centers of the future.

T-Systems relies both on its own data centers, like the one in Biere, and on operation in colocation mode. This means that the infrastructure and services are provided by T-Systems, while the data center is rented from another operator. Both have advantages. Colocation excels with fast availability and flexibility. On the other hand, having your own data center gives you much more freedom – e.g., with regard to the sustainability of the facility. T-Systems operates and uses a total of 33 data centers worldwide, 7 in its own operation, 26 in colocation mode. Today, the data center in Biere offers 20 large IT rooms with space for 100,000 servers and an IT capacity of 18 megawatts.

100 days

Before the data center in Biere went into operation, there was a 100-day load and complex test phase. During this period, all conceivable operating and load conditions, the functionality and resilience of the facility were examined. Huge load banks represent fully utilized IT rooms and prompt maximum demands on power supply and air conditioning technology. To simulate a complete outage, the testers then interrupt the power supply and check whether operations continue smoothly via the emergency power generators. In such a case, the data center in Biere can operate independently for at least 110 hours.

Tier III

The data center in Biere received Tier III certification from the Uptime Institute in summer last year. This confirms that the data center's plans are designed to provide users with a highly resilient infrastructure critical for business. And certifies the data center's availability and reliability.

Zero outage strategy

With its TÜV-certified Zero Outage quality program, T-Systems ensures maximum fail-safety of the entire IT infrastructure - the most important feature for products such as the Open Telekom Cloud. State-of-the-art security systems throughout the site protect against unauthorized entry, while complex encryption technologies protect against unauthorized access. All measures are tested on an ongoing basis to ensure that the data center, infrastructure and data are armed against cyber attacks or scenarios such as power outages. To this end, for example, transmission routes and communication channels are checked and how the supply of diesel is ensured for emergency power generators in different scenarios is looked at.

The PUE (power usage effectiveness value) in Biere is 1.28. Basically, the closer the value is to 1, the more efficient the energy use is. The methodology was originally created to be able to check at regular intervals whether the energy efficiency in a data center is improving or deteriorating. It is only suitable to a limited extent for comparing data centers with one another, because there are important aspects that influence sustainability but are not included in the evaluation. For example, the high availability required by customers and climatic conditions of the location are not taken into account. In addition to the PUE value, many other factors play a role in comprehensively assessing the sustainability of a data center – such as water consumption or the proportion of renewable energies.

There are many ways to further improve the energy efficiency of a modern data centre. What we are doing to become more sustainable:

Twin-core technology and emergency power supply provide fail-safety in data centers, air conditioning prevents servers from overheating and supports failure-free operation. These are necessary measures, but they require a lot of energy and have a negative impact on the climate balance of the facility and the PUE value. However, these precautions are necessary to ensure the security and availability of critical business data and services of cloud users. Outages and downtime cost companies a lot of time, money and customer satisfaction. Even further reaching consequences would occur for operators of critical infrastructure - and consequently also for society. Balancing this conflict of goals between fail-safety and sustainability as well as possible is our main focus today and also further into the future.

• Demand for computing resources is growing. However, the need for data center space will grow more slowly. This is because devices are becoming smaller and more powerful, which enables them to deliver ever higher computing output.  
• Demetropolization: Metropolitan areas will no longer be chosen as new data center locations. It will be much more important for renewable energy facilities such as wind or photovoltaic farms or consumers of waste heat to be in the area and, for example, for it to be possible for energy storage for hydrogen or batteries to be installed on campus.
• The usage scenarios will increasingly be in the area of multi-cloud environments. The sovereign cloud will become increasingly important for enterprises and organizations.