IT savings made simple
the day. Many IT organizations will have to make do with tighter budgets and leaner staffs—without neglecting critical IT functions. In fact, IT’s role may become even more important, helping the business respond to changing conditions and compete effectively for a larger share of shrinking markets. First, you must find a way to cut costs and do more with less.
In this series of articles, Citrix explores cost-saving strategies for every part of your IT environment, from the datacenter to the desktop. While the immediate goal is limiting IT’s impact on the bottom line, these strategies will also deliver meaningful long-term benefits by creating an agile, scalable infrastructure able to support future business needs, an essential IT priority in any economy.
Part one: Rethink your datacenter
As the most capital-intensive, power-hungry corner of IT, the datacenter is an obvious place to start looking for savings—but bring a scalpel, not an axe. Blunt-force tactics like deferred maintenance or delayed purchases can be time bombs in the long run, leading to costly and disruptive breakdowns of overburdened equipment. In good times or in bad, IT must preserve its ability to deliver reliable performance and meet fast-changing business requirements.
Virtualization offers a better approach. By consolidating datacenter services on a more efficient group of virtualized servers, you can reduce server count and realize immediate savings in equipment, administration, power and overhead—without sacrificing capacity or performance. At the same time, a static, complex datacenter becomes more dynamic, easier to manage and more agile in delivering workloads as needed.
Many existing servers are either underutilized—leaving valuable resources unused—or over utilized—threatening stability, maintenance problems and disruptions to the business. By virtualizing servers, you can balance virtual machines more evenly to ensure optimal utilization and significantly reduce the total number of servers needed—so that end-of-life machines need not be replaced and planned hardware upgrades can be completed at a lower total cost. You can reduce operating costs through faster application distribution, simple management and accelerated application delivery. Disaster Recovery (DR) becomes simpler and less labor-intensive as well: by leveraging the dynamic nature of virtual machines as secondary servers, you eliminate the need to run tests on parallel hardware, and you’ll have less hardware sitting idle at the DR site.
Explosive data growth is a key driver of rising IT costs and one of the trickiest to address. You can’t ask the business to reduce the volume of data generated by users and customers, but you can find ways to store it more efficiently using fewer resources.
Server virtualization offers some relief. You can run multiple servers with the same operating system drive and reduce storage across other virtual machines. Virtualizing storage itself delivers even greater savings. On a hardware level, replace direct-attached storage with a more scalable, efficient storage array network that offers capacity wherever it’s needed and minimizes unused capacity. Deduplication (dedupe) technology can dramatically reduce the total amount of data that needs to be stored. File-level dedupe is good, eliminating redundant attachments on mail servers and multiple copies of broadly distributed documents. Block-level and bit-level dedupe are better still, replacing repetitive elements within files with pointers. Think of how many large identical OS images, graphics and applications are currently cluttering your servers. By reclaiming this capacity, you can eliminate servers or avoid adding new ones as your data grows, and the savings are compounded by smaller, faster backups.
Virtualizing servers and storage will significantly reduce your power consumption and the generation of heat in your datacenter. As a rule of thumb, each 100 watt decrease in direct server power consumption yields an additional 20 to 30 watts of savings in your cooling load. You can achieve additional gains by using load balancing to direct users away from low-usage servers at certain times of the day so that they can be taken offline. Additionally, don’t overlook the low-tech low-hanging fruit: a $20 Kill-a-Watt device can measure the power consumption of virtually any appliance in your datacenter and help you identify gross offenders to replace with more efficient alternatives. While you’re at it, take the gadget with you on a walking tour of your users’ workstations. You’ll be amazed how even small savings can add up.
Not every cost reduction strategy starts with advanced technology. Even in a down real estate market, no company can afford to squander floor space. The virtualization of your servers and storage reduces the physical footprint of your datacenter. You can go a step further and optimize its layout to make it more cooling-efficient. Make sure your cooling ducts are aligned to support an optimal flow of cool air into the front of your servers and hot air out the back. Where servers have been taken out of commission, use blanking panels to ensure you don’t waste energy cooling an empty space. Check the Seasonal Energy Efficiency Rating (SEER) of your air conditioners, too; 18 is the most energy-efficient.
Measures like these make sense in any economy, both for the financial and ecological benefits they yield, and for the more efficient, scalable infrastructure they provide for the business. With money being tight, they’re now a powerful way for IT to demonstrate its creativity in responding to an urgent business need—and to deliver a real bottom-line impact without asking for sacrifices from the business in return. Along the way, you’re also improving fault tolerance, recoverability and performance, making your own life easier.