Inaugural Uptime Institute Server Roundup Contest

8

Howdy partners. We’re pleased to announce the first Uptime Institute Server Roundup. A roundup, according to Merriam-Webster, is the act or process of collecting animals (cattle) by riding around them and driving them in; a gathering in of scattered persons or things.

So get out from behind that desk and saddle up. It’s time to bring in those little doggies holed up in a server closet, sitting idle on the raised floor, wasting power and cooling resources. Based upon our field experience, up to 10% of those servers in your racks aren’t doing anything.

We recently heard of a major enterprise data center in New Zealand, fine cattle country by the way, that removed 20% of its servers that were sitting comatose.

Decommissioning a single 1U rack server can result in $500 per year in energy savings, an additional $500 in operating system licenses, and $1,500 in hardware maintenance costs. That’s not chump change.

Instead of piddling around shaving a point off your PUE, it’s time to get focused on what will make a real difference. It takes hard work to get in there and look for duplicated or unused applications, out-of-date, obsolete servers. We need you to drive in those wobbly old steer and replace them with more efficient, virtualized hardware.

Uptime Institute is offering a reward (or a bounty) on your obsolete machines. All participants in the Uptime Institute Server Roundup will receive a T-shirt and a badge for their web sites. Winners in each category will receive a commemorative rodeo belt buckle to proudly display on their corporate trophy rooms, desks—or if so bold—on their belts. The winners will also receive free passes and a dedicated presentation slot at Uptime Institute Symposium, May 13-17, 2012 in Santa Clara, Calif.

The rules: Two winners will be determined, one for most IT equipment removed, one for largest percentage of IT equipment removed. We don’t care how you get there. Going virtual? Server consolidation? Moving to the cloud? Going out of business?! We don’t care. Just unplug and decommission those machines.

Paperwork: What’s the proof? We want to see a paper trail. Send us change records. Do you Identify machines by server name or serial number. Removed 752 servers? We want to see the submittal of the work. Did you send the hardware to a recycler? Send us the receipt.

Results: We want to know how much energy you saved. Send us the UPS output reading before the change and after the change. You can do it right in the flow of work.

Photos: Send us a few before and after photos. Servers in the cabinets, servers in the docks going out. It doesn’t have to be exhaustive. Extra credit for creativity.

Submit your documents for inspection to Uptime Institute’s Hank Seader ([email protected]). Also send any comments, questions, cowboy jokes. Thanks for your interest. Hike up your Wranglers, unhitch your horses, and get to work.

Deadline: Contest closes Feb 1st, 2012.


Posted by mstansberry on 10-10-2011
Categories: Uncategorized
Tags: , , ,
 

Comments (8)

I think this is great way to make what is less than glamorous, FUN.

This is an excellent collective first step to retire inefficient and outdated servers. With Moore’s Law clearly still in effect, and with server manufacturers now focused on improving energy efficiency as well as performance, the key measure of energy efficiency in a server – transactions per watt – has been doubling every 18-24 months. Imagine the additional capacity this exercise could create on a global scale. By replacing older servers sooner rather than later, data centers all over the world can dramatically extend their useful life far into the future simply by following the technology development curve. It’s a lot cheaper to replace servers than to build new data centers.

A large enterprise can only survive by doing extensive lifecycle management if it does not want to build new data centers all the time. So be prepared to receive large numbers of decommissioned servers, if those companies also participate. However, as long as a company grows, the servers will grow on the virtual side and this also has an effect on the physical hardware. The second issue is the growth in kW on the storage side. It is more difficult to stop this and outreaches by far the growth in kW of the servers.

A very clever way to generate some fresh interest in this topic. Each inefficient system we upgrade puts us a step closer to our green dreams!

The research is clear: If you have servers that are older than two years, then from a TCO standpoint, mostly via energy savings, a refresh will most likely pay for itself. However, most organizations will get the first big “bump” in savings with very likely no capital expense and very little operating costs by taking on a project to remove all zombie servers. BRING OUT YOUR DEAD!

Excited to see the results. When will the winners be announced?

Look for an announcement in early March.
-Matt

Write a comment