If You Can't Measure It, You Can't Manage It!
Gone are the days of a year or so ago where businesses were frivolously splashing the cash on anything from office parties to new offices and bonuses. Even IT was making the most of the spending game, pouring a considerable amount of money into servers, storage and other infrastructure without any real consideration or planning.
But, organic growth and the legacy of the good old days has left a path of complexity with IT investments stockpiled, grossly under-utilised assets, siloed applications and increased energy consumption. Today's IT infrastructures, generally speaking are simply hot buildings full of underutilised servers - a real mess! Server sprawl means that IT managers are generally unaware of the exact assets they have on the floor, or even how they are being used. In fact, industry figures show that servers are currently only facing 20 to 30 percent utilisation rates, so there is a lot of room for improvement.
On the bright side, with some dedicated work, this is a part of the business that could see real cost saving, while still delivering the required service levels. To achieve maximum savings and to be able to successfully meet business needs, it is essential that IT has an accurate measure of its assets.
Management consultant, Peter Drucker famously once said; 'If you can't measure something, you can't manage it.' It stands true across all areas of business from supply chain management through to the data centre and IT infrastructure management. It is an old business adage, but one that is perhaps more poignant today than ever before. A marked increase in the focus on saving cost, providing an agile service to the business and operating in a more environmentally friendly way, means that the data centre must go through a dramatic transition. The idea is that to be able to meet these tighter business objectives, a complete infrastructure audit, inventory, and planning analysis to help predict growth and future IT requirements needs to happen before the infrastructure can become manageable.
Take storage for example. Think about how difficult it is to measure and plan around storage provisioning, capacity planning and service levels if there is no clear understanding of what's already there, and what's needed. Not only does this cause big headaches, but it also means that securing any sort of ROI can be problematic.
Yet the measuring of current assets can be labour intensive because of ever increasing data volumes and the 'silo' approach many businesses have taken in the past. So how can businesses make any headway in the mammoth task that lies ahead when it is often difficult to know where to begin?
Getting The Ducks In a Row
Some businesses will source an external consultancy to assess what is already being used and what is available and make recommendations to implement and predict changes. Determining business needs and impact accurately should ensure the right level of ROI is achieved. Some will try to unravel the puzzle themselves, in-house. Whatever happens to the data centre, it's all about making it easier and more cost effective to deploy and manage new technologies and serve the business.
IT managers must consider a number of questions: Where are we as an organisation? What assets are on the floor and how are these being utilised at the moment? Where can efficiencies be made? What are the future demands of the business now and in the medium and long term future? What are the costs of service currently? How are services and technologies measured, reported and managed?
The aim is to develop a foundation for a strategic, long-lasting and manageable infrastructure. Bringing this information together will enable organisations to streamline the data centre, optimise efficiencies, reduce the number of servers and ensure they are utilised efficiently. It will also allow you to accurately measure project successes and how they fair on the ongoing service level management the business requires.
Optimise the Data Centre
Once the analysis has been completed, optimising the data centre through strategic movement of assets and adopting new technologies can bring significant benefits. Virtualisation projects can be implemented, which mean fewer servers occupying less space, using less energy at lower costs whilst processing more data. Today, just four to five per cent of servers in the data centre are actually virtualised, so there is scope to reap further benefits from this technology. Data centre optimisation can also lead to a dramatic decrease in the amount of carbon emissions, and therefore a financial saving.
Data centre analysis can offer an insight into what inefficiencies may be causing greater emissions as well as providing a current state analysis of the environment. It can then provide recommendations on how to achieve a more efficient, greener IT environment based on industry metrics.
A recent server virtualisation project for a major pharmaceutical company saw the consolidation of over 1,400 servers down to just 700, reducing power and cooling costs by over 40%. The resulting carbon footprint impact is equivalent to taking over 4,000 cars off the road, or planting 2.1 million trees.
There are a number of other technologies and strategies that businesses can put in place to help consolidate and optimise their data centres. They need to create a list of the tools and technologies that can help meet objectives - such as tiered storage, de-duplication, cloud services and so on, and calculate the costs of implementing these technologies against the cost of not implementing them before taking the correct action.
IT departments understand that to be able to operate effectively on an ongoing basis, the tight management and monitoring of the data centre is crucial. Good management of IT infrastructure requires specialist skills, and when a job has never been subjected to the right degree of scrutiny, businesses need the best expertise available.
Short term solutions to these tasks don't come cheap, and skilled personnel are often hard to find. Businesses are experiencing increasing costs of recruiting and retaining specific skills. In practice most are unable to hire the necessary expertise at short notice to cover essential management tasks such as backup, virtualisation and security.
This means businesses are asking themselves if their IT spend is being used as effectively as possible, and indeed, are they best positioned to provide this IT service to their business? The answer is likely to be no, so the industry is seeing a shift in operation models towards the expert support model, by using off site services (OSS). This model involves handing over the management of key services to trusted expert third parties who deliver a quality service at a lower cost than the in-house alternative.
The OSS providers have already invested heavily in specialist consultants who have the expertise and time to dedicate to managing client infrastructure and can act as a complementary service to the existing IT department. This means that the ongoing management of the data centre is in hand while IT managers themselves are able to focus budgets on strategic projects that will save them money in the short and medium term. OSS also makes it easier for businesses to ready themselves for unpredictable requirements such as a sudden need for increased storage capacity or a project that needs dedicated IT staff for example.
This model is gaining pace. Gartner has predicted that outsourcing of infrastructure, applications and business processes will all increase over the course of this year and the early part of 2010. The analyst firm reports that spending on IT services grew by 8.2 per cent from £463bn in 2007 to £501bn last year. This year, it is expecting to see this figure grow significantly as businesses streamline their operations and outsource essential tasks.
There's no doubt that the data centre is in for dramatic change and its long-term future is anyone's guess. But one thing is for sure, the 'measure to manage it' model is essential if IT wants to succeed in saving money and serving the business in the best way possible.
David Leyland, UK OSS Business Manager, GlassHouse Technologies (UK) Ltd.
(ITadviser, Issue 59, Autumn 2009)