May 20, 2010 By Matt Williams
Springtime is a good time for housecleaning, as that well worn adage goes. The same can be said for IT, according to CDW-G, which recently released a checklist of suggestions for optimizing data centers.
Jerome Cheng, storage specialist at CDW-G, said state and local governments in particular are most interested in taking a periodic look at ways to become more cost and energy efficient.
And in many cases, that means taking an annual look at consolidation to avoid server sprawl: "A lot of small towns have anywhere from five to 10 servers, so they still can consolidate that. Instead of having five physical servers, you consolidate that into a single server, so obviously one server drawing power powers a lot less than five or 10," Cheng said.
Storage should also be examined yearly. "As state and local governments grow, storage requirements grow. Regardless of the industry, storage is always something that is going to be continually increasing," Cheng said. "Especially with state and local customers, they have a lot of mandates for archiving and retaining data for multiple years, and sometimes even for a lifetime."
Using a simple checklist can be a good idea because in this economic climate, many IT shops can't afford to proactively examine how to make their data centers more efficient. "A lot of the folks that we work with have a lot on their plate," Cheng said. "They are typically managing multiple systems, or worrying about several different projects that they are working on, with a limited staff too."
Consolidate data center server and storage systems -- Eliminating excess servers and storage equipment -- or even entire data centers -- can reduce energy and management costs. Blade servers can pack more computing power into less space, while server and storage virtualization can help allocate consolidated computing resources effectively, reducing excess capacity -- and therefore costs.
Update, replace or simply remove software and hardware that are no longer supported or way out of date -- It is one thing to be frugal about replacements and upgrades, but organizations can take on high operational and financial risk by running systems so far past their prime that little or no support is available when they break down.
Review desktop computing for opportunities to improve energy efficiency -- Opportunities to save significant money may be staring us right in the face: "Standing load" from unused computers or printers still plugged in; desktop computers and peripherals running around the clock when they don't need to be; failure to make the most of the power management functions built into desktop operating systems. For selected operations, some organizations could consider thin client architecture, which saves energy and can also reduce application support costs and boost security.
Ditto for the data center -- Power and cooling technologies have improved significantly, and blade server deployments tend to increase power and cooling requirements. If an organization has deployed new server and storage systems but still leans on the old power, cooling and management strategies, there are still more energy efficiency opportunities in the data center.
Make a tiered storage plan and make smart use of old systems -- Reduce, reuse and recycle. Match the investment in storage systems with the value and currency of the data residing on them, and deduplicate archived data. For long-term storage of inactive data or just for an economical, periodic backup of current data, even old tape systems can still work well in many situations.
Capitalize on cloud and hosted/managed services offerings -- For select software applications or hardware infrastructure, cloud computing and hosted or managed services can offer economies of scale, quality of support and convenience that many organizations would be challenged to match with internal resources. Outsourcing applications or facilities helps organizations declutter data centers.
Streamline operations by updating security group policy and user group designs -- Network updates can be painfully slow when one administrator is responsible for updating multiple groups and becomes a bottleneck. Streamline operations and free up time on both ends by updating the group policy design to grant access and rights to trusted users within major groups. Just be conscious that granting too many people access and rights to groups increases risk of data loss, so aim for an organized policy that balances productivity and security concerns.
You may use or reference this story with attribution and a link to