The importance of edge data centres for higher education
Download a PDF of this article Data centres a...
Data centres and networking are at the heart of the modern university. Effective IT planning and wise investment in infrastructure can help universities to address many of the challenges they face. Adopting an edge-based approach means many institutions can deliver better services to students and staff, helping them to stay ahead of the global competition.
The global nature of today’s higher education (HE) market puts pressure on institutions to deliver an even better working environment for their staff and students. The rise of Big Data and High Process Computing across all fields of research are increasing the demands on creaking IT infrastructure. There are also physical challenges such as space constraints and energy efficiency commitments. However, the most important issue faced by HE leaders is managing budgets in a sustainable way to meet changes in the education market. You can read the source Jisc blog here.
Though universities have traditionally had on-premises data centres, the sector could benefit significantly from following the trend towards the edge. With increased availability of cloud services specifically designed for HE, IT departments can refocus their limited resources to where they will have the most benefit.
Universities have traditionally operated a centralised IT infrastructure for data and compute functions. With the introduction of education-specific co-location and shared data centres by Jisc and Janet in the UK, the options for university infrastructure design and development have changed considerably.
Just as content providers are shifting capability away from the cloud to the edge of their networks on a micro scale, so universities can benefit from the same approach. This may be of particular benefit to multi-campus and city-based universities.
The cloud offers universities an opportunity to move some of their less sensitive data and non-critical systems to an off-site infrastructure. This can reduce costs and may be particularly beneficial where space is at a premium. Conversely, data and systems that are mission critical can be located directly adjacent to where they are required using microdata centres at the edge of the university network.
Smaller, decentralised data centres at the edge of university networks can provide a range of benefits, including:
A larger number of smaller data centres can also simplify the design and implementation of new infrastructure by reducing the power and cooling requirements in a single area.
The availability of all-in-one cabinets, such as the Rittal Micro Data Centre, makes this an increasingly viable option. Built around a standard rack with integrated environmental control, power protection, security, and fire detection and suppression systems, these allow the deployment of a micro data centre almost anywhere. Larger scale ‘off-the-shelf’ solutions can be deployed quickly and easily to provide departmental or campus-level data centres without the complex design overheads.
The recent Jisc HE Sector Barometer Results, which surveyed HE leaders, identified six key transformational challenges that university IT teams face:
Source: Jisc HE Sector Barometer Results, November 2017
Organisational change will be key to meeting these challenges. However, moving away from a centralised approach to IT infrastructure towards a more edge-based data centre strategy can support universities in delivering what is required.
The Jisc survey found that supporting a truly agile organisation is the challenge that HE leaders feel requires the most effort.
Traditional data centres have complex design and build requirements, which can take 12–18 months to deploy. Coupled with the extended procurement procedures that universities typically have to follow, it is not possible to deliver new or enhanced data infrastructure in an agile way within existing budgets.
Edge data centres – in particular, modular microdata centres – are quick to deploy and scalable. These lend themselves to an agile approach, allowing IT teams to respond quickly to organisational changes and requirements.
UK universities are now competing in a global education marketplace. Rankings, research and funding rely on HE institutions recruiting the best and brightest from around the world. Increasing the number of overseas students can provide a healthy boost to universities’ balance sheets, as well as creating a diverse learning environment. However, in order to do so, universities must compete against world-leading institutions – some of which have much deeper pockets.
Today’s students have grown up in an always-on digital world. As such, the shift towards digital learning technologies and the use of rich media is changing the nature of university education and the requirements it places on IT infrastructure.
E-learning platforms, blended learning and the flipped classroom, where students prepare for lessons in advance at home, are becoming central to many university courses. As such, students have come to expect fast, easy access to their online study resources in halls of residence, the library and across the wider campus. As the quality of these resources improves and the bandwidth requirement increases accordingly, edge data centres can help to take the data where the students are. For fragmented campuses and those where real estate is at a premium, microdata centres can provide a useful means of meeting changing student expectations. This approach mirrors that taken by content providers such as Netflix, who in the battle against latency and bandwidth costs have been moving their content out to the edge of the network.
As e-learning becomes an integral part of teaching, the technology demands within the classroom will also increase. From high-definition lecture capture to digital resources and ubiquitous Wi-Fi, the demands on infrastructure have increased exponentially.
To ensure universities are delivering the best learning experience, IT infrastructure needs to support high bandwidth, low latency and, most importantly, easy and reliable access to digital resources. Edge data centres can help to simplify the network infrastructure, making delivery, maintenance and access management easier. The approach also supports high-bandwidth applications by reducing the need for cross-network data transfer. The modular nature of the small cabinets used for edge implementations allows for a great deal of flexibility in their location and application. This allows IT teams to meet changing requirements more quickly and at a lower cost.
Large centralised, multi-purpose data centres can pose challenges when it comes to access management and data security. Edge data centres provide a clear separation between different data sources and can simplify user access management for different resources. Taking a more decentralised approach to HE infrastructure that includes cloud services, central on-premises provision and edge can help to keep sensitive data safe and secure, while making it easier to access shared data.
Modular, micro data centres are available with inbuilt security and monitoring systems, including intrusion detection and biometric access control. This allows such systems to be used in any location regardless of the sensitivity of the data. Integral fire suppression, water and dust-proofing options also allow edge data centres to provide protection against physical risks in harsh environments.
Container-type data centres such as the Rittal RiMatrix S can provide an off-the-shelf backup and mirroring option for disaster recovery. These self-contained, fully featured data centres can hold up to eight 42U racks. With power, cooling and security built in, these data centres can be located anywhere.
In addition to supporting universities in meeting specific challenges, edge applications can offer advantages in all aspects of data and compute usage in HE. However, there are some applications to which edge data centres are particularly suited.
Many research-led universities rely on high-performance computing to keep their research at the cutting edge. While some applications will always require large-scale clusters, other research dependent on compute capability can benefit from small-scale implementations.
With the advent of high-density, high-efficiency compute platforms, it is perfectly feasible to place one HPC cabinet directly in the lab for research requiring dedicated compute capabilities or with a need for low latency.
Research is creating more data than ever before. Providing storage close to the data source can prevent losses through transmission errors or hardware failures, as well as providing a route to greater redundancy and safer backup procedures.
In research, the data collection and analysis phases may be distinct. A modular, edge approach is scalable and allows compute capability to be easily included, or kept separate, as per the requirements of the project. Self-contained microdata centres also enable a modular approach, allowing additional units to be added as project requirements (or budgets) increase.
As more collaborative projects between different departments and institutions are launched, the data infrastructure requirements shift. Jisc’s shared data centres provide an unparalleled opportunity for shared research, particularly where large data sets are in use.
These facilities allow HE institutions to develop a micro data centre outside of their own IT infrastructure. They provide the flexibility for institutions to use their own hardware and give direct access to the fast Janet Network while offering shared maintenance and support services. This approach can simplify the management and security overheads for joint projects, making collaboration easier.
As the internet of things (IoT) and smart buildings become more widespread, university estates are likely to see an increasing requirement for connectivity. Building information management systems (BIMS) can generate large quantities of data and the analysis can require significant compute capability.
Universities frequently have some of the most ambitious environmental strategies, but face the challenge of implementing these within an old (or even ancient) physical infrastructure. IoT and BIMS may provide a solution to meeting environmental goals, but their implementation may be dependent on IT provision. Dedicated micro data centres within buildings can reduce the impact on the network that hundreds of additional connected devices might have. They would also reduce latency for real-time systems, allowing for improved efficiency and greater gains.
The self-contained and scalable nature of many products designed for edge applications is ideal for implementation on university and other HE campuses, with small micro data centres, data storage and connectivity that can be installed almost anywhere. For more resource-intensive requirements, pre-fabricated data centres can be quickly and easily installed and even offer containerised versions. Rittal’s range of edge data centre products are ideal for the university market, with simple configuration tools, flexible applications and cost-effective pricing.
Find out more about Rittal’s range of edge-ready cabinets and pre-fabricated data centres, all available from UPS.
If you have any questions about edge data centres please contact us here.