The Dangers of ‘Just In-Case’ Capacity Planning
In a frenzied example of ‘just in-case’ planning, anxious shoppers across Briton have been turning to stockpiling over the last few months as the answer to Brexit uncertainty. Now with no deal all but off the table, much of this ‘worst case scenario’ food could go to waste.
Seem extreme? Perhaps. But businesses are falling into the exact same ‘just in-case’ trap when it comes to capacity planning. And while everyday consumers arguably have little-to-no control over the Brexit outcome, businesses do have the ability to mitigate such uncertainty and cut costs on capacity spending by up to 35% in the process.
Planning for Complexity
When planning an estate, firms have a lot to consider: the network, the infrastructure, the applications, inside the firewall and synthetic monitoring from the internet. That’s a lot of technology that needs to be working round-the-clock, meaning 24/7 monitoring via a range of tools is a necessity to build operational resilience and ensure the whole estate is performing correctly.
What isn’t a necessity, however, is the over-provisioned capacity that many firms factor in to cushion themselves against systems failures. The typical enterprise customer over-specifies the hardware it needs when it deploys an application, leading to an average over-provisioning of more than 30%.
At a basic level of avoiding the worst case scenario, over-provisioning makes sense – if the exact behaviour of the estate is not known, it is necessary to err on the high side. However, this over-capacity costs both capital spend on hardware, and operating costs on software, power and people to run the estate. In the cloud alone, provisioning a hosting application just one size bigger than what’s actually needed roughly doubles the cost, equating to $65 billion a year in wasted cloud spend.
Getting to Know Your Estate
By collecting the data from the monitoring tools, and using the latest Capacity Planning Tools, it is easy to find out the correct size that a business estate should be. Through building application demand models, the estate can be ‘right-sized’ for the peak load that it has to support, or even sized for a peak demand expected in the future, all while avoiding both wastage and system failure.
Recovering the over-provisioned estate may take time, and revolve around smarter replacement of legacy equipment, or a move to virtualisation or the cloud. But with huge savings on the table, firms should have no qualms in opening their arms to capacity planning.
At the end of the day, when it comes to business capacity, uncertainty is a choice, not a reality. By capturing monitoring data in a repository and analysing it with the latest capacity planning tools, any firm can easily right size its estate. And unlike a cupboard full of pasta and canned goods, it is a guaranteed return on investment.