The amount of data that businesses generate has led to difficulty in building and maintaining on-site data centres. In response, many organisations have shrunk their investment in in-house infrastructure and IT workforces and moved instead to the cloud — with an abundance of choice service types, models and vendors, there are fewer barriers to adoption than ever.1
For some, however, retaining a local model is necessary to meet specific needs (such as security for organisations dealing with highly sensitive information). The appeal of the cloud that other organisations have benefited from is tantalisingly out of reach. However, there is an option for those looking to keep a close handle on their data while reducing their costs that is worth considering: colocation.
Why colocate?
Where there are no readily available, mature cloud services to suit an organisation’s particular needs, or for those who aren’t quite ready to start their journey into the cloud, there is still an immediate and pragmatic step to move away from in-house data management. Colocation refers to a situation where a business privately-owned servers and networking equipment are physically hosted by a third-party data centre. This is often a simpler first step towards using a third party service for external hosting; colocation introduces a shared, or reduced responsibility on the organisation to host its own workload, instead moving its servers, hosting computers, and data offsite.
In this scenario, the hosting data centre is responsible for the physical security of an organisation’s equipment and supplying it with power. The business, on the other hand, takes responsibility for its servers, computer processing, storage, virtualisation, data and applications.
Colocation may be a good solution for businesses facing the following challenges:
- The workload type is unique such as with legacy or non-supported IT systems or where a unique licensing or support arrangement exists with a vendor or software product.
- The workload volume is significant, for example, where the large volumes of data are collected as a part of normal business processes (such as in banking institutions) would incur huge costs to store or move.
- The information classification of the work doesn’t suit cloud storage. For example, health and law enforcement agencies deal with private, sensitive and top secret data and are restricted from adopting cloud services because the data must remain within a designated region to comply with data regulations.
- The proximity requirement to related technology is high. For example, where an organisations’ interrelated devices must transfer data at high speed over reliable connections, for instance SCADA Systems (supervisory control systems) or involving real-time operational machinery, such as in industrial manufacturing.
Where colocation shines
Data centre services have matured greatly in recent years, and can automate and scale almost all of their services, reducing downtime and increasing power usage efficiencies. Data-centres-as-a-service (DCaaS) are going beyond their core remit to suit their clients’ varied needs, including offering pay-as-you use models, uptime of 99.995%, dual power, backed by carbon neutral efficiencies, and quick access to technological innovations.
Major data centre players have regionally dedicated low-latency data centre networks (DCNs) powered by optical and fibre technologies that connect centres and regions. Enterprise grade DCaaS often have their own networks or operate in partnership with interconnected data centres around the world, and are backed by high speed network connections with reduced latency and improved performance specifically suited to serving globally dispersed clients and end users. As 5G and IoT services mature, businesses are using colocation services and creating ‘edge data centres’ — smaller centres located close to the businesses/populations they support — using colocation services to provide quick geolocation services to customers.
Cloud services are essentially a collection of data centres located all over the world and the top three providers account for approximately two-thirds of all rentable data service computing services.2 In fact, most cloud providers use a colocation model to host their own workload, as their core business is providing cloud services not managing data centres. As a result, data centres can offer their own colocation customers a direct connection to the top public cloud providers such as Google Cloud Platform, Amazon Web Services (AWS) and Microsoft Azure which they also host. Colocation customers therefore get the opportunity to integrate their infrastructure and cloud services seamlessly to create custom hybrid cloud solutions.
Choosing the right provider
Colocation comes with its own nuances that need to be carefully managed. Ownership, contracts and costs should be scrutinised before signing on to any data centre’s service. Importantly, the provider must be backed by reliable investors and have innovation and security at the forefront of their business strategy.
Given all organisations have unique requirements, and considering the wide variety of data hosting options to choose from, how can businesses ensure they make the right decision on where to put their data? In essence, they will need to have:
- A clear strategy: Whether organisations are hosting their workloads themselves, colocating or using a hybrid model, it must be done with a clear hosting strategy. This is vital, as it will set out the terms of engagement with business and enterprise partners and outline the technical architecture of how systems can connect and share information. Just as in private enterprise, government agencies — with constantly changing customer service needs and ever-evolving information systems — will need a strategy underpinning their technology. Government procurement of data centre spaces and services are based on asset refreshment cycles, the end of outsourcing contracts, the end of life for the data centre, or requirements to expand data centre capacity securely.4 DCaaS providers are quickly adapting to these requirements by offering secure regional colocation services, and going the additional mile of certifying their services are fit to host classified, secret and top secret data.5
- A transition without inhibition: Storing data securely in a data centre should be accompanied by a strong understanding of the data distribution, the variety and volume of data being collected, stored, and processed; and how data can be made accessible to customers and partners. Usage should also be accompanied by a fail-proof exit strategy which ensures that data and computing power can be moved from one technology, system or data centre provider to another when a contract expires or needs change. Exploring options to enable this by virtualising servers or containerising workloads — arranging applications and processes into self-contained, moveable packages — is a sound strategic investment.
A cloud-led future
Innovation in data centres and DCaaS is evolving rapidly. Organisations exploring solutions for their data storage should ensure they put an emphasis on revisiting, revising and gaining insight into the role data plays (or should play) in their future business strategy. This will inform decision making with regards to data hosting and the possible usage of cloud services in the longer term. In turn, this will help direct a company’s IT strategy, paving the way towards adopting and integrating the technology and processes which will ultimately enable transformation and encourage growth.
