There is a lifecycle for every piece of equipment in the IT arsenal. From servers to switches, hubs to printers, and even networking cables; all IT equipment must be retired and replaced at some point. The question is, how long should the equipment be in service? What do we do with the equipment once we decide to remove it from production? If it is still running well, should we retire the device entirely?
Each IT solution and situation should be looked at independently with a few rules of thumb. If we are too quick to decide to fully retire a piece of equipment, it may end up in a landfill, long before the end of its usable life. E-waste has a detrimental impact on the environment, but our equipment recycling decisions may be able to slow that destruction. And when it finally comes down to the decision to get rid of the equipment, there are some very compelling landfill alternatives which I will describe later on.
First, when considering the piece of equipment in question, there are a few points to look at before considering its retirement:
- Is the piece of equipment mission critical?
- Does it have any redundancies built in?
- Is it clustered (paired redundantly) with one or more of its kind?
There are of course other questions that can be asked to help narrow down the decision, but I find these are a good starting point. I have used these questions at HFR for many years, and they have helped drive our networking budget for years, saving us significantly on the bottom line.
Mission critical? Yes, big question. In the IT world, we have all dealt with that server that, if it went down, it actually would not have a massive impact. We are so often preoccupied with achieving six sigma optimizations, that we forget that sometimes it is not really the end of the world when a server goes down. If the given system yields a convenience, but is not heavily relied upon, maybe there is a workaround that can be put into place while the system is being repaired. For example, we use a fax server to cut down on paper usage—if it failed, we could easily use a fax machine while we repair the server. That fax server had no redundancies and was fairly old, but it handled the workload perfectly, so we used it for a non–mission critical system. This way of thinking/planning can help avoid the early retirement of a decent machine that is perfect for its purpose.
Built-in redundancies? Well, did you pony up the extra bucks for redundant power supplies? Did you skimp on the additional drive cage that would yield a better, redundant disk solution? Unfortunately, IT isn’t always given the best budget for a situation, and sometimes we can’t afford the Cadillac. I have had to face the scary situation where I had no choice but to put a piece of mission critical equipment into service that only had one power supply. Other times, only a single disk! Yikes! But if we invest in a machine has some redundancies built in, it can potentially stay in service a bit longer than its non-redundant counterpart.
Is it clustered? Clusters of equipment have long been a holy grail for IT teams longing for an uninterrupted good night’s sleep. When a machine goes down that is part of a cluster, the other machine(s) pick up the slack and continue operation for system. So, even if a singular machine does not have redundant parts of its own, it has counterparts to rely on, should it fail. In a scenario such as this, especially when there are three or more machines/devices in a cluster, you have a lot more breathing room when it comes to older equipment.
Time to Retire
In our network, we have a backup site that serves as our disaster recovery solution. We have found the ideal balance of maintaining healthy equipment lifecycles while reducing costs by sending older devices off to an easier, less stressful life. If the device still has some good, usable life, it may find a second life in our datacenter, as part of our disaster recovery solution.
When we are ready to retire a piece of equipment, we look at its capabilities and decide if it is a better machine than the ones already in our datacenter. If the machine is newer, it provides an opportunity to replace an even older piece of equipment. But there are also times when an older device is more powerful than the newer machine, so further scrutiny is always required. We have created several clusters in our disaster recovery site with old production units that are still quite powerful. So, even though the machines have some age and may fail, no worries, because they are clustered! Another option is to target the oversaturated used parts market; you may be able to upgrade a machine for an insanely low cost—the perfect solution for a disaster recovery site.
Before you go tossing out a perfectly functioning machine because it is several years old, consider some alternatives. Perhaps it still has value in your network, or maybe there are other options before sending it to the landfill.
Consider the Environment
In the past, when a machine was going to be eliminated, there was a good chance it was just going to end up in a landfill somewhere. IT equipment is highly toxic to the environment, so it is irresponsible not to explore landfill alternatives. At HFR, for over 17 years, we have sent our used equipment off to a refurbisher. The refurbisher cleans up the equipment, makes any necessary repairs, then donates it to organizations that cannot afford new equipment of their own. It is an incredible, win-win solution, both socially responsible and environmentally friendly.
We live in a throw-away world. Our landfills are filling up with junk that will be around for hundreds of years, rotting away and leaching chemicals into the soil and water. There are times when a piece of equipment just does not belong in your network any longer. Instead of pitching it, consider donating it. Take the tax break and feel good about helping someone else as well as the environment.
Ask yourself the three questions we pondered. Is it: (1) Mission critical? (2) Built with redundancies? (3) Clustered? We can draw useful conclusions about the usable lifetime of our equipment by asking them. With out-of-the-box thinking and careful planning, you can get the most out of your equipment. And when the time comes to retire the machines, they can have a second life elsewhere.
A downloadable copy of the article can be found here.
Contact the Author: Bradley Bergstrom, CTO | firstname.lastname@example.org