Data Centers

How to make the most of outdated servers, storage, and switches

There are times when you really do need the best available technology. But there are also times when you're wasting money on unnecessary upgrades.

gorodenkoffistock-899720520.jpg
Image: iStock/gorodenkoff

Show me any technology department with a newly reduced budget, and I'll show you a sysadmin who found clever ways to keep outdated equipment in service.

There are plenty of applications that really do need the latest-generation hardware—you don't want to run a transactional database on something from 2015—but in most other cases it's fair to wonder if vendor recommendations for refreshing systems every couple of years may be more salesmanship than engineering.

SEE: Computer hardware depreciation calculator (Tech Pro Research)

"That's conditioned behavior," joked Mark Nichols, chief product officer at Curvature, a secondhand IT provider in Charlotte, NC. "If they sit down and think about it, there's only a portion of their environment that really needs the latest and greatest."

"Our value proposition is to stretch that from 3 years to 10, or longer," Nichols explained. Especially for non-critical, non-virtualized applications running Unix, Linux, AIX, and so on, "I think there's still a lot of bare metal out there... you've got a lot of juice in that lemon."

SEE: Tech budgets 2019: A CXO's guide (ZDNet special report) | Download the report as a PDF (TechRepublic)

Other times, if you have a cluster of older machines, perhaps they're not individually up to modern standards, but it can be cheaper and faster to expand the cluster with another older machine rather than replacing the whole setup with a single new beast, Nichols noted.

Alternately, you can always repurpose older generations of servers, storage, and switches; secondary and remote offices can be good places for that equipment in cases where capacity and speed are less important. "Then you can run it for a second life, another 5 years in that secondary capacity, and lower your operating costs as well as your cap-ex cost... I think naturally IT departments do it this way," Nichols said.

Outdated systems can always receive interim upgrades such as solid-state drives, broader network cards, and NVMe memory. Eventually, you will reach processor limits, although peripherals tend to be the tougher bottlenecks.

SEE: How to calculate depreciation on computer hardware: A cheat sheet (TechRepublic)

Yet another option is to build your own systems. Curvature can replicate some types of Intel-based network switches, or you can have an intern learn on the job by assembling an open-source Backblaze network-attached storage (NAS) unit. (Backblaze is known for its reference designs that are a powerful compromise between a commercial NAS vs. a homemade or no-name JBOD chassis.)

It's important to never decide on repurposing older equipment without understanding the risks and other issues. Curvature publishes a refreshingly honest blog about this: Recent posts include New Tech is Buggy Tech, Do You Really Know Your IT Life Cycle?, The SSD Blind Taste Test, How to Stay on Budget and Innovate, and How to Avoid the Biggest Rip-Off in Networking.

The recent Gartner IT Sourcing, Procurement, Vendor & Asset Management Summit in Orlando offered five sessions about IT disposition. Nichols presented a session with Sean Zongker, a procurement director at Dell—illustrating that even the top companies selling new hardware are keen about making the most of their internal systems.

Also see

About Evan Koblentz

Evan became a technology reporter during the dot-com boom of the late 1990s. He published a book, "Abacus to smartphone: The evolution of mobile and portable computers" in 2015 and is executive director of Vintage Computer Federation, a 501(c)3 non-p...

Editor's Picks

Free Newsletters, In your Inbox