What bothers data center providers most? The industry research from Prylada
December 15, 2022
Prylada has conducted a series of customer development interviews with experts from the data center industry. We also interviewed several representatives of vendor companies, which provide services to data centers and are familiar with their processes. We tried to collect as much valuable information as possible to properly assess what is going on there inside.
In this article, you will find direct quotes of the experts, summary infographic, main conclusions about the challenges the industry faces nowadays, and the issues that come first. We dive deep into the topic to see how we can better help in resolving the data center problems with Prylada. That’s why we pay so much attention to every tiny pain data center providers may have.
First, a general interview infographic
The ever-increasing demand for storage, processing, and efficient control of a large amount of information drives the constant growth of data centers. Designed to coordinate company’s vital IT operations and assets, data centers are expected to ensure a high level of security and robustness. Still, there are challenges which make data center owners stay concerned and exert a lot of effort to increase the data center competitiveness and efficiency.
In this article, we gathered the most common challenges the interviewees shared with us. The points that we heard more frequently are placed first. This does not necessarily mean that they are the most critical ones, but it does indicate their prevalence. So let’s get started.
Challenge #1 Power efficiency and sustainability
Servers and devices demand electricity, but incorrect distribution of energy may result in power outages. The current tendency towards server consolidation and virtualization helps reduce hardware in data centers, but this does not always reduce energy consumption. Server consolidation typically refers to the situations when standard rack servers are being replaced with blade servers.
Reference: As Google says, a blade server is a stripped-down server computer with a modular design optimized to minimize the use of physical space and energy. In general, a blade server consists of a chassis, or box-like structure, housing multiple thin, modular electronic circuit boards, known as server blades. They are called blades because of their ultra-thin shape. Each blade contains a single server, often dedicated to a single application.
At the same time, data centers must be available when their customers need them. This includes having redundant power and cooling systems in place to ensure that the data center can stay online even if there is a problem with one of the systems.
Importance of geographical location
Many data centers are dependent on the standard power infrastructure, which might be problematic in terms of availability and position. Therefore, the location of a data center should be near an area with a lot of reliable power sources, so it can keep running without any interruptions in case there's a power outage or disaster in the area.
Moreover, a favorable location of a data center in terms of climate conditions becomes a significant advantage, which helps reduce energy consumption and costs. Colder climate and windy weather provide natural cooling for data center equipment. That’s why more and more data center providers are selecting northern countries for their new buildings. Additionally, data centers can use renewable energy sources available in the region, such as solar or wind power, to help offset their energy usage.
Backup power to restore servers
As equipment requirements vary, power and cooling specifications are becoming increasingly crucial. Especially with the global energy crisis, when it is very difficult to provide a constant source of energy to sustain a datacenter, unless renewable resources are considered.
Every data center should get backup power, which might be a powerful turbine or an alternative energy source. Backup power does not need to run the whole facility for a lengthy amount of time. It only has to be on long enough yet to safely shut the system off until the project’s overall power is restored.
Installing a dedicated power source for data centers and server rooms is a good place to start. However, in order to be better prepared, such solutions must be maintained. It is critical to restore servers as fast as possible, which necessitates improving the boot procedure.
Data centers must provide a high level of security for the equipment that is stored within them. This includes protecting against physical threats, such as theft or vandalism, as well as cyber threats. Customers place their trust in data centers to ensure their information is protected.
Protection of personal data
One of the biggest challenges data center providers are experiencing in 2022 is meeting compliance around protecting individual data.
Organizations with physical locations that collect data on visitors, contractors, and employees who sign in to the workplace collect sensitive data. As such, they have a responsibility to protect the personal data of each individual through encryption, software, and processes. Some of the aspects data centers need to take into account include:
- How they collect and store personal data
- How they provide informed consent to collect data
- How they protect the data of individuals.
Currently, many organizations have their physical facilities and online presence out of date to meet privacy needs. To avoid penalization and fines, it’s important to update workflows, cookie policies, and use encrypted software and hardware for data collection and storage.
Data center security is dedicated to safeguarding access to private and sensitive data. It helps attain access when the necessary data is misplaced. However, digital content and data are disturbed by outages, ransomware, and corrupted information.
Physical data center security
Numerous cybersecurity solutions incorporate real-time monitoring of data channels, but physical security requires the same attention. Data centers require protection from people with malicious intentions, intruders, and prospective disasters which include water leaks, fires, and cooling system failures. Every security door has to be locked properly, and access should only be given to authenticated personnel. Still, to be extra cautious, their exit and entry must always be logged.
Cameras should be installed in the server room and throughout the facility, with security personnel constantly monitoring the streams. This helps the security team to act swiftly upon identifying an unauthorized individual. Additionally, every employee within the server room should also be monitored. If a data center does not have or permit cameras in the server room, they should instead point them at every entrance and exit. The security personnel can then watch who enters and exits the facilities and examine what equipment is carried into and out of the rooms.
Challenge #3 Environmental issues and server cooling, and how they are connected
Data centers not only use a lot of energy, but they also create massive amounts of heat. This requires an energy-intensive cooling system, which is, taking into account the skyrocketing energy costs, becoming increasingly expensive. And with temperature increasing globally each year, this quickly becomes a vicious cycle. With data centers constantly growing in demand, and environmental issues worsening, data center owners and managers will need to adapt properly in order to maintain performance.
When talking about hyperscalers, they are built to go fast. To prevent overheating and loss of data, they're cooled in a very specialized way using a large amount of water. And then they need to have stable power and abundant amounts of power. So, sometimes they have to put legacy fuels in the back of the data center.
It also depends on how cooling is done. To reduce negative impact on the environment, more data centers are moving to or trying to move to oil cooling versus water in air conditioning style systems.
The pressure to stop using water for cooling servers also contributes to the rapid development of liquid cooling technology.
Challenge #4 Balancing cost-cutting measures with efficiency
In order to remain competitive, data centers need to be able to keep their costs down. This includes both the initial cost of setting up the data center and the ongoing costs of running it. At the same time, to keep up with the demand for data, data centers are using more and more power, and providers need to find new ways to manage their increasing energy costs.
So how to balance cost-cutting measures and efficiency? Budgeting and cost management are ongoing concerns for all departments, but the data center's circumstances are unique. While you want to ensure that your data centers are efficient, secure, competitive, and beautiful, you must also keep costs in mind.
Traditionally, customers sign multi-year deals with the big hyperscalers and those facilities are designed to be very flexible in the sense that you can do any type of computing. So it's a general purpose of a data center. And so the companies pay for high availability at those facilities 24/7. They pay for the ability to put different types of compute, real-time compute, and non-real time compute in the facility.
There's a cost associated with that, and lots of IT organizations are beginning to realize that they don't actually use all of those capabilities, even though they're paying for them. That's starting to create the concept of a multi-cloud, where you can sign up with several different providers for different purposes. And one of those purposes might be to take really specific jobs and move them to a platform that is tuned for that job. And because it's tuned and that's all it does, it's a lot cheaper.
Challenge #5 Old data equipment
Many data centers are looking to save money by potentially buying old data equipment. While this can save money, it might not be the best decision in the long run.
Every new generation of equipment features a higher level of security and efficiency, which the previous generation can no longer compete with. Here’s the reason why data centers are recommended to replace their equipment every 2 to 4 years. For this changeover period, new equipment is released, and upgrades become necessary.
Data centers that ignore replacing equipment with new models, or even worse - step down to the previous releases, are likely to face efficiency and security problems, which in turn may affect operation of the entire data center. Additionally, older equipment consumes much more energy than the modern generation, which correlates with the next challenge.
The pandemic became another bottleneck for equipment modernization, which negatively effected the supply chain and caused severe damage to the data center sector. The building of new facilities has been delayed, and the refresh cycles of data centers have been disrupted by labor scarcity and a lack of equipment supplies. To meet the challenge, businesses currently invest in equipment far in advance and use mergers and acquisitions.
Challenge #6 Capacity management
A data center management is difficult. Identifying and removing risk from data center operations is of the utmost importance in order to achieve high availability for customers and high reliability for important systems.
Because data centers have a lot going on inside, unexpected failures are unavoidable. Applications, connecting cables, network connectivity, cooling systems, power distribution, storage units, and much more – are all active at the same time. Maintaining peak performance necessitates running the data center at total capacity. Nonetheless, IT managers frequently leave a margin for error, a capacity protection gap, to ensure that activities are not disrupted. As a result, resources and space are wasted. In addition, power and energy are being wasted.
A data center infrastructure monitoring (DCIM) system provides more detailed information about data center operations and performance metrics. It allows managers to track, analyze, and generate reports in real-time, allowing them to make informed decisions and take immediate action.
The Internet of Things (IoT) is another possibility to remotely control the digital and physical infrastructure of a data center in real-time. On the other hand, according to Gartner, the Internet of Things may turn into a disruptive force that will revolutionize the data center due to the enormous amount of data it will generate. The information collected by IoT devices still needs to be sorted, prioritized, stored, and evaluated.
Challenge #7 Facility management and scalability
Anyone who has worked in a data center for more than a day is familiar with cable mess. Getting rid of existing messes is difficult. It might not seem so difficult to run a 50ft CAT6 cable to your network infrastructure at first glance. However, you soon find yourself surrounded by cables and unable to recover. Therefore, every rack should be equipped with patch panels to minimize cable mess, which requires additional efforts.
At that, data centers need to be able to quickly scale their services to meet the demands of their customers. This includes being able to add more capacity and space when needed, and providing the necessary support for new applications and services.
Another facility issue refers to “blind utilities control”. Companies that rent facilities for their data centers usually don’t have access to the real values of how much electricity, water, or other resource has been consumed in a certain month. They receive common bills that are hard to validate. Such companies have to find solutions for additional level of utilities control.
Challenge #8 Transfer to edge computing
Controlling the amount of data is another big challenge faced by the industry these days. This is a challenge because controlling huge chunks of data requires better technology, such as edge computing.
Reference: Edge computing is a networking philosophy focused on bringing computing as close to the source of data as possible in order to reduce latency and bandwidth use. Smaller centers are being built regionally as cybersecurity teams try to keep up with the new challenges presented by this advancement. In addition, organizations are requiring less cloud resources from their providers and those data centers are not being used fully. So, it becomes important to increase the bandwidth reception of data centers. Additionally, providers will need to find ways to reduce latency and improve network performance.
Data centers are under pressure to improve energy efficiency and provide a higher level of security for co-located equipment. Security and privacy measures need to be put into place to make sure that no one can access any sensitive information on the servers inside the data center. In addition, data centers need to be able to scale quickly to meet the demands of their customers.
Still, some of the challenges listed in the article might be predicted, as the main concerns of the data center providers are frequently raised on different platforms, including social media. However, during our personal communication with the experts, we discovered a lot of things that were new to us and made us start thinking from other perspectives. The collected information will definitely help us adapt Prylada to fulfill the customers’ needs even better. We hope the article was also useful for you.
And finally, as solution always starts with a problem, it’s easily define what industry trends to expect in the following years. So, keeping all the challenges in mind, what are your bets?
We’d like to thank everyone who participated in our Customer Development Interview:
- Stavros Zavrakas, Founder of Orthogonality
- Boris Jabes, CEO & Co-Founder of Census
- Michael McNerney, VP of Marketing and Network Security at Supermicro Computing
- Isla Sibanda, Cybersecurity Specialist
- Piotr Marciniak, PhD , CEO & partner of Telco companies
- John Belizaire, CEO at Soluna Computing
- John McDonough, Sales Operations at EXA Infrastructure
- And other industry professionals, whose comments were taken for this article