Although it wasn't recognised as such, the first data centre was conceptualised soon after the dawn of the computer age, when systems were complex and characterised by chaotic cable runs and disorganised environments.
With the 'client-server' revolution of the 1990s, the computer room was the logical place to store the growing number of file servers and storage devices common to most forward-thinking businesses. New standards for structured cabling allowed the computer room to eventually 'morph' into the 'data centre'.
Today, the data centre is defined as the central repository for the storage, management and dissemination of information associated with an organisation via its networks and underpinned by its applications. It may be a physical or virtual entity.
In the data centre, virtualisation is essentially the process of separating the hardware and software elements, and then distributing the data infrastructure across multiple independent sites that could conceivably be located across town, in other cities or anywhere on the planet. Virtualisation is about how computing is done, not where computing is done.
Virtual necessity
Sometimes referred to as the abstraction of computer resources, virtualisation is essential to a cloud computing deployment. Employing cloud technology, organisations are able to do more with available resources while rapidly adapting to changing business dynamics and new workloads.
Virtualisation gives the data centre manager the ability to pool and dynamically allocate disparate IT resources across any number of business units, allowing services to be quickly deployed and even scaled to meet growing needs in moments. It underpins the philosophy of the cloud-supported data centre, which relies heavily on the sharing of resources to achieve the numerous benefits associated with economies of scale - including greater efficiencies and optimised productivity.
For example, organisations are continually demanding more bandwidth. It remains an important quest, supporting the diverse set of applications now becoming available - particularly in the BYOD (bring your own device) space - and will surely impact the future design of the cloud-supported data centre as well as the management of the corporate network, from edge-to-core.
To meet this and other challenges, the design of a cloud-supporting data centre requires skills as diversified as an appreciation of evolving malware threats to a clear understanding of - and appreciation for - latest 'green' standards.
Clear vision
It also demands a holistic, 360-degree view of the facility's migration path for its major elements: servers, storage, networking and applications to support a continuously expanding user base and deliver a user experience that keeps pace with escalating information processing demands.
This ability is key to the adoption of appropriate architectures that will allow a smooth uptake of future technology-based opportunities.
Virtualisation is about how computing is done, not where computing is done.
Martin May is regional director at Enterasys Networks.
These objectives need to be targeted whether an organisation is looking to adopt a private cloud (to serve users in its own organisation), a public cloud (to deliver services to a broad user population), or a hybrid cloud that is essentially a blend of the two technologies.
According to Ignacio Llorente, head of the research group at Complutense University, in Madrid, future enterprise data centres will look like private clouds supporting a flexible and agile execution of virtualised services, combining local with public cloud-based infrastructures to enable highly scalable hosting environments.
Against this backdrop, cloud-supporting data centre designers must face up to one of the important challenges of the coming age - the very real possibility of a future deterioration in network predictability and visibility, as virtualisation deployments are dramatically scaled up and out, and immense numbers of local and storage area networks (SANs and LANs) are unified and fully automated.
For IT management to guarantee consistent service provisioning and quality of service in such an environment, attention must be paid to the tools needed to enable end-to-end automation across the entire corporate network.
Designers will need to find a platform for reducing network tiers, flattening data centre networks and creating more intelligent network infrastructures that minimise latency and maximise capacity and throughput.
Importantly, they will need to introduce streamlined management facilities enabling them to automate provisioning for users, devices and applications while reducing the time it takes to deploy and support new systems and new-generation learning tools - which will surely proliferate even further in the cloud computing environment of tomorrow.
Share
