Frequently Asked Questions

Data Centres

A data centre is a high availability facility that provides the space, power and cooling required by computer systems, data storage devices and telecommunications systems. Data centres need extensive infrastructure to operate, including redundant power systems, uninterruptible power supplies (UPS), backup power solutions such as generators, air conditioning, fire suppression systems and various security systems. These facilities are vital for the smooth functioning of various organisations and the wider digital economy, as they provide a safe, secure environment for data and computing equipment, with constant monitoring to ensure optimal performance.

The purpose of data centres is to store, process and disseminate data and applications. Every day, companies need computing equipment to run their business applications and activities, such as email and files sharing, communication and collaborations services, productivity applications, CRM, databases and more. Data centres support the provision of important IT services such as data storage, backup and recovery, data management, networking, and distributing large amounts of data.

Data centres come in several types, each designed for a specific kind of need. The main types are:

• Enterprise data centres: Also called on-premises data centres, these are owned and operated by the company they serve. They are typically located on the company’s site.
• Colocation data centres: Here, a company rents space within a third-party data centre to host its own servers and other hardware. These facilities are also known as multi-tenant data centres.
• Cloud data centres: Cloud data centres are the places that cloud infrastructure is housed to support the provision of large-scale cloud and digital services online. Some are owned by the major cloud providers (examples might include Google, Amazon and Microsoft) and some are provided by third-party operators. These facilities are sometimes known as hyperscale data centres due to their large scale capacity.
• Edge data centres: These are smaller facilities located closer to the populations they serve, designed to deliver cloud computing resources and cached content to end-users with less latency.

A colocation (colo) data centre is a facility where multiple organisations can lease space for their own servers and other computing hardware. In a colo, the data centre operator provides the power, space, cooling and physical security, whilst granting access to third-party carriers to provide connectivity. This enables organisations to benefit from the economies of scale associated with large data centres — including redundant power and cooling resources, and high-speed network access — without having to manage those resources themselves.

Data centre connectivity refers to the networking and interconnection methods used within and between data centres. This includes everything from the basic physical infrastructure — such as routers, switches and fibre-optic cables — to the protocols and services that allow for data transmission. Good connectivity is essential for ensuring fast, reliable access to data and services.

A carrier-neutral data centre, also known as a network-neutral data centre, is independent of network providers. Since it’s not owned by a single Internet Service Provider (ISP), multiple telecom carriers can access the facility, allowing interconnectivity between them and/or colocation providers. Customers can choose from a wide range of services without being locked into a single carrier. This fosters competition and can help drive down costs and increase service quality for customers.

A peering point, also known as an internet exchange point (IXP), is a physical location where networks come together to peer or exchange traffic between their networks. Peering points help reduce costs, improve speed and increase redundancy by allowing data to take more direct routes between networks.

Data centre security involves various measures and technologies designed to protect a data centre from threats – both physical and digital. Physical security measures include access controls, surveillance systems and secure facility design. Digital security measures can include firewalls, intrusion detection systems, network segmentation and encryption. The goal is to protect the data stored in the facility from unauthorized access, as well as protect the infrastructure from disruption or damage.

Private cloud refers to cloud environments dedicated to a single end user or group where the environment usually runs behind that user or group’s firewall. Public cloud is the standard cloud computing model where service providers make resources, such as applications and storage, available to the general public over the internet. Hybrid cloud is a solution that combines a private cloud with one or more public cloud services, with proprietary software enabling communication between each distinct service.

Cloud computing is a service that delivers shared computing resources, software or data as a service across a network. With cloud services, you do not need to physically maintain or manage servers or storage: servers are owned by the cloud provider and data is managed virtually. Colocation, on the other hand, involves customers leasing physical space within a data centre to house their own hardware. Customers maintain and manage their servers, storage and network, while the colocation provider takes care of the building and its utilities.

A cloud gateway is a network node connecting two networks that use different network protocols. It acts as a bridge between a local environment (like a data centre or office) and a cloud service provider. Cloud gateways enable businesses to securely connect their existing networks to cloud services, facilitating data transfer.

London is becoming constrained, so out of town data centres can offer scalability with easier access to land and power, can be in lower risk geographies and can potentially have lower costs depending on the market or geography.

Data centre sustainability involves designing and operating data centres in a way that reduces their environmental impact. This involves a range of strategies, such as using energy-efficient hardware, improving cooling efficiency, powering facilities with renewable energy, and responsibly disposing of or recycling old equipment.

PUE is a metric used to determine the energy efficiency of a data centre. It is a ratio of the total amount of power consumed by a data centre to the power consumed by the IT equipment within it. The closer a PUE value is to 1.0, the more efficient a data centre is considered to be.

HPC involves the use of supercomputers and parallel processing techniques to solve complex computational problems. HPC systems have the ability to process data at high speeds and handle large data volumes. They’re often used in fields for simulation and modelling in diverse applications, such as climate and weather modelling, quantum mechanics, bioinformatics, precision medicine, financial risk assessment, fraud detection, computational fluid dynamics and more. One of the primary applications of HPC is Artificial Intelligence (AI).

Data centres require power for running servers, cooling systems, storage systems, networking equipment, backup systems, security systems and lighting. They are powered by a combination of power supply distribution, backup systems and management tools to ensure continuous operation. Data centres typically receive power from the utility grid through multiple power feeds which is distributed to IT equipment through a series of power systems: transformers, Power Distribution Unit (PDU), Uninterruptible Power Supply (UPS) and switches. Redundancy and backup systems are essential to minimise downtime and maintain business continuity. They can include UPS, which provides immediate short-term power in case of a grid outage, and generators, which can provide longer-term back-up power. Increasingly, data centres are also making use of renewable energy sources to reduce their environmental impact.

Power density refers to the amount of power kilowatts used by each cabinet that houses servers. According to the Uptime Institute, average power densities hover around 8-10 kW per cabinet (or rack). 40 kW to 125 kW is considered to be high-density, but some extreme-density racks can consume up to 200 kW. The more power per rack, the higher the computing workload that can be accommodated using less floor space. HPC, AI, machine learning (ML), internet of things (IoT) devices and big data are some of the major contributors to higher-density demands.

Data centre cooling is essential to keep servers and other hardware at optimal temperatures to ensure they run efficiently and safely. IT equipment is heat sensitive and has to be operated within strict environmental specifications. There are several methods of cooling, including air-cooled systems (using Computer Room Air Conditioning (CRAC) or Computer Room Air Handler units (CRAH) and chillers), liquid cooling, water cooling and containment strategies. The goal is to remove the heat produced by the equipment and keep the temperature and humidity levels within a range that is safe for the equipment.

In a colocation data centre, services typically include the leasing of space (which can range from a rack to a cage or a dedicated room), power, cooling, physical security and network connectivity. Additional services can include remote hands (onsite technical support), physical storage, logistics, high security fitments and more. The specific offerings can vary by provider and are often customisable to meet the unique needs of each customer.

When selecting a colocation data centre, consider factors such as the location (for latency, disaster risk and legal compliance reasons), physical and digital security measures, reliability (uptime), scalability (availability of additional space as your needs grow), power density, connectivity and the level of support services available. It’s also important to assess the reputation and financial stability of the provider.