|
For over a decade, our data centers have been in the grip of a crisis caused by lack of IT transparency, and it's getting worse. The crisis is fuelled by companies' increasing need for the latest technology to support their critical business applications, creating a relentless cycle in which the pace of change is forcing infrastructure teams into a permanently reactive mode. In truth, many large organizations are not good at keeping track of what they already own, and how it is being used to support the business.
Due to this lack of visibility, the governance teams in control of new project funds can find it easier to buy new equipment than to share or redeploy what the business already has – idle or underused – elsewhere.
At the core of the crisis is a fundamental reversal in the economics of provisioning large-scale IT infrastructures. Floor space and environmental conditioning now cost an order of magnitude more than the computers themselves. New business initiatives can easily justify the capital expenditure to buy more hardware, but they rarely acknowledge the effect this has on longterm operational costs.
Space and time
Every new generation of technology means machines that are faster, have a higher power density, and occupy a smaller space footprint. This all adds up to extra wattage, extra heat to get rid of and an IT estate in which older, less powerful machines occupy far more physical space than their contribution to the business now warrants.
Brute-force remedies, like increasing power and cooling facilities, have a limited shelf-life, and even major decisions – like the multimillion dollar decision to build new data centers – can be derailed. For example, electricity and real estate supply limitations mean that the construction of new data centers in Manhattan, NY and the Docklands area of London is severely restricted. A senior IT executive working at a prominent global financial institution comments: “We have a massive amount of square footage, but at the same time we don't have the right space, at the right wattage and cooling capabilities in the right places around the world. Our data center situation is strangling our business agenda”.
For many organizations that rely on low network latency – like financial institutions that decompose transactions for distributed execution – wholesale relocation becomes an extremely complex option, even if the sizable budgets required are available in principle, as computing resource is required as close as possible to the trading floors.
A way of continuously optimizing the datacenter is required. This needs to start with an understanding of how each server is used, which business applications it supports, and therefore what the value to the business actually is. In this way unused, unnecessary and inefficient hardware can be weeded out or refreshed, and software consolidated.
Visibility
It's hard to understand what's actually going on in our data centers – complex crossdependencies between systems, the diversity of stakeholders, and the cultural distance between IT and the overall business all make for an uncomfortable reality.
In data centers around the world, managers really don't fully understand what's there and what it's all doing. This results in inefficient data center resource planning – where far more infrastructure than is necessary is employed to meet business objectives, and far more time than is necessary is spent trying to meet SLAs. The processes that give us visibility of the hardware, software and systems across our IT estates are ripe for an overhaul – they are manual at best, and due to the high rate of change, always produce an out-of-date view. We can only begin to clean up the mess in our data centers when we have a clear picture of IT assets and dependencies across the organization.
Towards a solution
Fortunately, our understanding of the dynamics within data centers is beginning to shed light on concrete techniques that can help turn the tide. Process support and sophisticated, new technology tools are combining to create an approach for data center optimization that improves clarity and visibility, reduces risk and offers attractive infrastructure benefits to many organizations – such as better business agility, lower operating costs, reduced risk and an ability to implement transparent and justifiable internal chargeback policies.
|
