At the end of a year and the beginning of the next one many companies and individuals make certain predictions. They might be in the field of economy, finance or science (such as the case of, for example, IBM and its 5 in 5 predictions). We monitor the predictions of experts regarding the telecommunication field. In fact, we ourselves can predict the enhancement of certain trends, considering that in a number of cases, everything is so obvious.
During the current and the previous month we have collected the opinions of several experts and analysts, adding our own view on the matter, and now, we can share all of it with you. Of course, analytics is a highly partial field, but some things are so obvious that it is impossible to be wrong. So, let us see what the New Year can bring to Data Center operators and the entire telecommunications sector.
More attention towards the security of Data Center infrastructure
Despite the fact that cyber security is becoming a more serious work area (when, in fact, was cyber security not a serious matter?), bad-intended minds continue to break the most, or so they seem, at least, secured systems. That is why currently companies are reinforcing this direction, trying not only to counteract the attacks from hackers, but also to recover from these attacks as soon as possible. The most important thing to many is to keep their infrastructure up and running, preventing it from failing, something that could make the data center go offline. And nobody is overreacting here- the danger is as real as it seems. According to a study conducted by the Ponemon Institute, last year about 22% of Data Centers, on which a hacker attack was carried out, went offline for a certain amount of time. This study was called Cost of Data Center Outages 2016.
Besides that, another problem is the recovery of the infrastructure after successful attacks. Surprisingly, most companies operating in the telecommunications sector, are not very well aware of what to do after an attack has already occurred. The majority of managers and employees of these companies consider an attack an unlikely event and are not preparing themselves to deal with its consequences.
Micro - Data Centers are become more and more relevant
Mobile solutions have always been relevant, but currently, with the expansion of the IoT area, such solutions have become truly irreplaceable. Besides IoT, the so-called Industrial Internet of Things (IIoT) is now developing rapidly, and the relevance of tendencies such as broadcast streaming of media content is increasing. All of this is helped to manage by micro - Data Centers, that can be positioned as close as possible to the infrastructure.
As a result, data collection and analysis (e.g. from various IoT servers) is significantly simplified. Micro - Data Center slightly differ from conventional data centers, the main difference being the size. Moreover, they are usually made modular to be quickly deployed, when needed. The task for the industry now is to create micro - Data Centers and the development of tools for their monitoring and control.
Micro - Data Centers, among other things, are usually equipped with specialized solutions for their infrastructure. These primarily concern power supply and cooling.
Reducing exploitation expenses
A very important topic for Data Center operators is the reduction of maintenance costs. Over the past few years a number of interesting solutions were created to reduce energy costs. Among other solutions it is worth mentioning the creation of intelligent cooling infrastructure management systems. These are, for example, free cooling and neural networks. Google itself managed to reduce the power consumption in its Data Centers by approximately 40% thanks to the use of the DeepMind inovation.
AI from DeepMind (the so-called weak form of AI) was trained to control the server hardware and infrastructure elements (e.g. the cooling systems). 120 different parameters of the DC activity were assigned for control, including control over the functioning of air conditioners, opening or closing the windows, fan speed and many others.
The algorithms create are self - learnable, and can operate in a variety of settings, including unplanned ones. And this is only the beginning. Currently DeepMind is trying to convince Google managers on the need to install additional sensors within the data centers, that will collect information for the control algorithms. Receiving more and more data on the state of the data center in real time, the AI will be able to manage its work more effectively.
This type of solutions will soon cease to be exotic and will become a regular practice. It will allow companies to reduce maintenance costs in a variety of expense areas, it is not only about the cooling.
DCIM - solutions
Generally, the term refers to the integration of the information technology and management functions for the centralization of control, management and activity planning of the main infrastructure systems of the data center. This type of solution is usually needed by the Data Center engineering team.
Who needs all of this? First of all, a successfully selected and customized DCIM-solution allows to improve operational efficiency of the entire data center, as well as of individual elements. As a result, the cost of IT-service drops when converting to one user per time unit. In the long term, it allows to improve the reliability of the Data Center activity while improving the services provided to users. Another pleasant side is the decrease in operating costs for Data Center maintenance.
For Data Center operators DCIM-solutions are needed to address the problems of different specificity. This could be the management of hybrid cloud platforms or DPC clients maintaining remote monitoring on their assets. There is a large number of DCIM-based schemes, and all of them can benefit the operator.
Further operations with HDD
Despite the fact that many DPC operators are switching to solid-state drives, a full transition to this alternative is not yet possible and is unlikely to be possible in the near future. The reason is the significant difference between the costs of SSD's, HDD's and hybrid drives (especially when it comes to professional equipment). So far, it is the hard drives that provide the best combination of capacity, operational speed, reliability and price.
Additionally, the HDD manufacturers and suppliers are releasing new hard drive models with an increased capacity and operational speed. It is very likely that this particular year enterprise HDD's with a storage capacity of 16 TB will be available on the market.
Switching to Lithium
Indeed, nowadays lead acid elements of valve-regulated (VRLA) battery chargers are very widely used. However, it is far from the best solution for accumulator systems. The problem is both the size and heat dissipation, and also the fact that VRLA require frequent maintenance works.
It's been a while since solutions based on lithium-ion batteries emerged, and these solutions are getting more accessible. They have several joint advantages, being smaller than VRLA's, therefore taking a smaller amount of space, making extra room below the server hardware other type of systems. Also, these type of systems require less maintenance work and produce less heat.
Indeed, modular schemes are getting more and more popular, as they have a great number of advantages, with only a few disadvantages. First of all, it concerns the possibility of a quick deployment of modular data centers, with simultaneous commissioning. Besides that, such DC’s are massively scalable and operate quite effectively. If needed, their equipment can be quickly changed or renewed, as these types of solutions are usually connected.
Generally, the above are the main ideas that we wanted to mention. It would be interesting to also find out your thoughts on what will Data Center operators encounter in the next year, which trends will prevail and which ones will do the opposite.