How a Server Finds New Life at ServerMonkey
Are you curious how ServerMonkey refurbishes servers? To give you an idea, we’ve decided to take you on a journey. We’re going to let you experience what it’s like for a data center server to go through the refurbishment process at ServerMonkey. After all, our refurbishment process gives you what you’re looking for—a server with a new lease on life.
Where it all begins
As you know, many organizations replace their servers every couple of years because, back in the day, older servers weren’t reliable and processor performance doubled every 24 months. But today, servers are more reliable, processor performance gains are slowing, and organizations, from Amazon on down1, are discovering that late-model servers have plenty of value.
ServerMonkey helps organizations leverage that value.
Decommissioning a Server
So imagine for a moment that we’re going to follow a 2U x86 server, used for virtualization, living in a data center someplace in Houston. Over the years, it has worked well, but today it’s being decommissioned. Virtual machines are migrated off, a remote power off command is sent, and an admin comes by with a server lifter, disconnects it from the network and PDUs, yanks it from the rack, pulls hard drives to be wiped or shredded, and then moves it to a pallet at a loading dock, alongside another dozen decommissioned servers.
That’s because ServerMonkey’s purchasing team is buying those servers.
A team from ServerMonkey shows up, with a truck, at the dock, and puts pallets of servers on the truck. Those servers take a journey to ServerMonkey where they show up at a locked, staged teardown area. A team member puts a blue cone on the pallet, which indicates that the pallet of servers needs to be checked into the system.
What might surprise you is that, at this point, our team doesn’t really know what we have on the pallet. Our Teardown team gets a list from the Purchasing Team, and some of our team members reconcile what server models are on the pallet with the list from Purchasing. We make sure we have what we paid for, and then it’s time for teardown.
We don’t just put up servers on the shelf. What we do instead is tear them down to component level so that we have server chassis and a stockpile of components, including processors, drives, network cards, and the like, ready for custom builds.
We may have more than a hundred servers sitting on pallets at a time, and we need to prioritize teardown, which we do based on inventory and customer requirements. The server we saw before is a late-model 2U rack server, and those are in demand. So the teardown team gets to work on it right away. We strip it down and put all its components into individual bins by part number. Separating components let us manage inventory, lets us set up online configuration tools, and lets us order extra components if needed to fulfill an order.
The server chassis, essentially just a case and motherboard, is then shelved until it’s needed.
When ServerMonkey gets an order, possibly on the phone or through an online configurator, it’s for a specific configuration. Whether it’s being configured as a storage server filled with many high capacity drives or a data analytics server with the fastest processors, ServerMonkey takes the order and assembles the server. We don’t stockpile various server configurations, anything we sell is built-to-order. Assembly also includes cleaning. We give each component a once over with contact cleaner, to make sure any contaminants are removed from electronic contacts.
Once the server is built, ServerMonkey makes sure it works. We conduct two kinds of testing, static and dynamic. In static testing, we update BIOS and firmware to make sure that everything is up to date, and configuration matches the new components. Then in dynamic testing, using built-in vendor-provided testing suites, we test every component, including memory, drives, and networking.
Then once the basic tests are completed, the server goes to quality control.
At this point, our quality control test confirms that the server assembly matches the order.
Then they conduct additional testing to make sure that the server is operational. If it matches the order and passes testing, it will be packed up and shipped to a customer. If a component fails additional testing, it will be replaced by the quality control team.
Once we’ve completed tests, we’re ready to ship the server out.
We wrap each server in a pink poly sheet, seal the top with an ESD label, and place it, with a start-up guide, TekShield pamphlet, and other items inside a box that’s prepared with a 100% filled foam in place bag. We include power cords, bubble wrapped rails, bubble wrapped bezels, and software licenses. At this point, we take a picture of all the parts and the unit inside the box for quality assurance.
Then we fill the box with foam-filled bags, at the four corners, along each side, and on the top. These bags are created by injecting them with a mixture of foams that expand and harden to conform around the server and keep it in place.
Finally, we seal the box with staples and tape, add a Fragile sticker, and we’re ready to add an address label and ship it to you.
Hope you enjoyed that journey through our process. As you can see, our approach to refurbishing servers means that our high quality refurbished servers deliver even higher value. Our focus on custom assembly, testing, and quality assurance ensures that customers like you stay satisfied. For more information, fill out this form!