Putting storage first

17 Oct 2012

Every time an organisation embarks on a new IT project, there is the temptation to buy ‘best of breed’ components to suit that project’s specific purpose.

All too often, storage is not considered in its own right; it’s simply assumed that the best option is to go all-new, with storage purchased on a per-use basis to fit a particular project or requirement.

The problem stems from the fact that, in the majority of cases, the decision to acquire storage infrastructure comes after the choice of server and application has been made – storage remains an afterthought. Even if IT managers do take the entire stack into consideration, the idea that storage implemented a few years previous could be suitable for the exciting new purpose is easily disregarded over concerns of speed bottlenecks and incompatibilities.

The main issue with such a project-based, multi-platform approach to IT procurement is that as needs change and storage technology develops over time, an organisation ends up with separate islands of disparate storage; using different protocols, disk types and, more importantly, different management standards. In a worst-case scenario, a company will end up using a small percentage of capacity on multiple storage systems with no ability to share disk space or even inter-communicate.

And yet, the negatives are considerable – capital outlay is multiplied, as acquiring a range of different storage platforms quickly becomes prohibitedly expensive. Multiple licenses also need to be purchased for different activities, from encryption to thin provisioning – adding to on-going costs. At the same time, data is growing faster than budget; storage requirements are growing up to 40%* a year in many organisations, and this percentage is increasing as the type of content stored – such as high-definition video - becomes more space-hungry.

When a company that has purchased a number of siloed storage systems outgrows one or more of them, its only option is to rip and replace at significant cost. Techniques such as de-duplication and compression will slow the tidal wave of data down, but can only ever be a temporary solution and come with their own performance pitfalls.

The only true way to ensure that storage is used effectively and efficiently throughout its lifecycle is for organisations to put storage first. Instead of designing the application tier, then the server and network tier and then finally thinking about the storage tier, storage should be the foundation upon which any infrastructure is built.

Employing the right method of storage is also a key consideration in efficiency. Think about how a supermarket stacks shelves; the most common items are usually stacked at eye level, with less common items stacked lower down or higher up. In storage terms, more commonly used data is stored on high-speed fibre channel storage, while less frequently accessed data can be stored on SATA or even tape devices. But again, Information Lifecycle Management (ILM) relies upon an integrated, and agile, storage foundation comprised of storage systems that work together, not against each other. Having to use the wrong method of storage because a particular storage appliance was never designed for the purpose you’re trying to use it for will ultimately lead to performance issues right through your organisation – the result being a loss of productivity and custom.

By putting storage first and ensuring a single storage foundation that can be easily deployed across not only current projects but also any new infrastructures that may be required in future, businesses can ensure that any influx in data can be handled effortlessly and in a highly cost-efficient way.

*Frost & Sullivan / IBM, March 2011

Join Chris Bone on Wednesday 17^th October at 11:50am in the Data Centre Optimisation theatre to hear how Fujitsu’s Agile Storage Foundation is putting storage – and customers’ needs – first.