Data Center Architecture: There is no doubt, in addition to people’s health, Covid-19 has also had a major impact on many sectors and companies. The impact that has been reflected, and will be reflected even more, also on the architecture of the data center. The widespread diffusion of smart working has led to a rapid decentralization of the IT infrastructure. In almost every industry, people have had to rely on cloud services provided by a data center to communicate, without forgetting the use of streaming, social media, and gaming. Thus, many business conversations have moved from meeting rooms to home studies.
Many companies have had to implement digitization plans or accelerate those already in place to address the situation. Several of the solutions created to deal with an emergency have had such positive feedback that they will be maintained, if not even strengthened, in the future. So, where previously data centers were driving the rise and growth of digital technologies, now the need for digital technologies is driving data center growth.
But how? What are the trends emerging for data center architecture?
Table of Contents
One aspect that emerged predominantly in 2020 is the need to have the certainty of business continuity, even in the face of exceptional catastrophic events. According to a new Future Scape report, by the end of 2023, this should lead 80% of companies to move to cloud-centric applications and infrastructures at twice their pre-pandemic speed. This a trend that will also be reflected in data centers, for which Gartner forecasts that in 2021 there will be a 6% increase in spending compared to 2020, reaching $200 billion.
The “advanced” use of virtualization in data center architecture allows for great efficiency and a high density of machines, an ideal solution to meet the redundancy and computational needs of rapidly growing companies that need great scalability. In all likelihood, this will primarily be helped by hyper scalability, a very effective way of meeting the growing use of digital services that generate huge amounts of data. But there is the flip side: hyper scalability is used in high-density racks requiring an adequate power supply and cooling system. Something that not all companies can achieve, both for the availability of the environment and for the costs.
An alternative is co-location, which companies increasingly use. This service involves renting the necessary space from specialized suppliers and obtaining the appropriate computing, network, memory, and storage capacity as and when needed. Thanks to new intelligent monitoring solutions, co-located systems allow CIOs direct control of all parameters of their IT systems at any time. In this sense, data centers can offer business intelligence software and monitoring tools on actual network usage patterns, with effective resource control.
Il cloud on premise
Among the advantages of co-location is the opportunity to have a pay-per-use service. Conversely, the data and applications are not premise but must be ported to the provider’s machines.
However, for reasons of latency, compliance, or the use of legacy applications, not all companies can afford to leave data and applications outside their perimeter. Despite this, they can still reap the benefits of the cloud by employing on-premise cloud service. In this case, the data center is still “in-house,” but of the available machines, only those that are needed are used, and the costs relate only to the actual use.
This way, if business needs require it, you can scale quickly without having to buy new hardware, and you don’t even have to worry about updating machines or security: it’s all the responsibility of the service provider.
The role of automation in data center architecture
The ability and willingness to customize are crucial when setting up a data center. The road that is being followed to address them is by choosing best-class solutions such as HPE GreenLake. This means that next-generation data center architecture includes AI-powered tools that will drive autonomous systems, causing them to take timely actions to prevent outages and ensure high availability. Intelligent monitoring systems, operating 24/7, will send alerts to IT managers if the pre-set data center usage thresholds are exceeded. The same goes for network security and performance.
One parameter that defines the data center’s capacity is the rack density, i.e., the number of servers that can be placed in a rack or in the entire data center. More servers mean more resources but also higher consumption. While manufacturers are doing everything they can to limit them as much as possible, IDC says the amount of energy used by data centers continues to double every four years. And data center energy use could exceed more than 10% of the global electricity supply by 2030.
To reduce energy consumption, the data center architecture, understood as the architecture of the building where the data center is located, plans to incorporate energy-saving systems and the possibility of exploiting alternative energy sources, such as solar turbines, wind, and gas.