Home       About Us      Media Coverage      Get More From Existing Equipment
Asset Maintenance & Virtualization Top Choices For Extending Equipment
Check out any data center that’s been online for a decade or two, and chances are there will be legacy equipment still humming along earning its keep through some sort of useful service. A few years ago, data center managers might have cringed at the thought of a 10-year-old server or tape backup system, but today, the ability to lengthen equipment deployments is the mark of a well-managed IT department. 

While the general rule for data center hardware has typically been three to five years between upgrades and replacements, in the current tight economic climate, organizations are extending life cycles in an effort to squeeze more ROI from existing assets, especially servers, storage, and power equipment. 

The two principal methods for stretching the life span of IT assets are proper maintenance and leveraging newer technologies, especially virtualization. 
Key Points
Establishing a comprehensive virtualization strategy that combines server and storage virtualization will help prevent expensive hardware purchases and maximize existing IT investments. 
Moving parts are subject to wear and tear. Keep air intakes free of dust and debris. 
Install regular service packs, upgrades, and patches to operating systems, firmware, and applications to ensure optimal performance. 
Excess heat can rapidly reduce the life expectancy of practically all data center assets.
Rob Enderle, president and principal analyst of the Enderle Group, explains that data centers should shift from policies of equipment replacement to those focused on preventive maintenance and system longevity. He notes, “This typically means making sure thermals stay toward the lower end of the recommended range.” 

Offering ideas on how to achieve Enderle’s advice, Gerry Cullen, founder of NetBotz (www.netbotz.com) and IT WatchDogs (512/257-1462; www.itwatchdogs.com), believes that investing in an automated environmental monitoring system is a small price to pay for data centers to ensure against potential heat spikes that could significantly reduce the lifespan of servers and other necessary networking components such as routers and switches. Cullen says, “Heat and humidity, even in slightly elevated amounts, can damage a variety of equipment.” 

“As to keeping older gear in play longer, it’s really a matter of blocking and tackling—knowing the workloads well, tuning and pruning as needed, being disciplined about how the systems are managed. It’s much like keeping an older car going for as long as possible. It’s not rocket science, it’s just careful, disciplined, economically minded maintenance and upkeep,” adds Jonathan Eunice, principal IT advisor for Illuminata. 
“Resources can also be more aggressively load balanced using tools like virtualization so you make the best use of capacity and [ensure] that loads are dynamically shifted to appropriate hardware based on need so you make the best use of the performance you already have,” Enderle adds. However, he cautions to be sure to include standby hosting because failures will come up with this model, and when load exceeds capacity, there will need to be alternative available resources. 

Eunice sees virtualization as the first and foremost tool one would use today to get longer usable life out of equipment. “Generally, though, servers should be spec’d for virtualization—plenty of processor cores, loads of memory, good I/O and storage subsystems,” he advises. 

But that strategy can require an investment cycle. “Despite the higher ROI, new gear can be a hard thing to justify. The same applies in storage, where features like thin provisioning, pooled storage capacity, IP-accessibility, and built-in deduplication make for long-lived investments—but if you don’t have those things now, it often requires an investment cycle to get there,” says Eunice, who echoed the same tips for facilities equipment including HVAC and power distribution. 
Storage Shelf Life 
Bob Fine, director of product marketing at Compellent (www.compellent.com), also cites virtualization as a way of extending asset life cycles, especially when it comes to storage. “By utilizing technologies such as virtualization and using storage equipment that can scale to accommodate future storage needs, customers can avoid expensive rip-and-replace upgrades and extend the life cycles of their existing assets.” 

Fine also suggests two options for stretching storage life cycles—automated tiered storage and SRM (storage resource management) applications. When implementing a tiered storage strategy, data centers keep mission-critical data on Fibre Channel drives and older data on more economical SATA drives. “By using more SATA drives and fewer high-performance Fibre Channel drives, users can extend the life of existing drives and hardware,” suggests Fine. 

Similarly, through the use of an advanced SRM application for monitoring and managing the storage environment, Fine points to a reduction in storage costs through simplified storage administration, a more streamlined disaster recovery strategy, and better reporting on storage usage and energy costs. 
Good Housekeeping 
Don’t overlook the importance of providing clean air for cooling all the equipment within the data center. A properly maintained HVAC system is also an integral part of boosting equipment life cycles. Filters should be inspected and changed on a regular basis to ensure the delivery of clean air, as well as maintain the efficiency of the HVAC system itself. 

Carol Blake, founder and CEO of Access Floor Systems (www.accessfloorsystems.com), explains that despite changing filters regularly, dust and debris can still gather in the raised floor plenum and air flow ducts. Blake concludes, “Keeping your data center clean and well-maintained will increase the overall useful life of electronic and mechanical equipment.”
by Sandra Kay Miller