Parts on the shelf: A waste or self-insurance?

Every IT environment will keep some amount of spare parts. But how much is too much? IT expert Rick Vanover shares some pros and cons of this practice.

I’ll admit it, I’ve hoarded parts. This can be things such as spare servers, hard drives, cables and other critical components of the data centers I have worked in over the years. I would venture to say that all of us in some capacity have held on to some parts over the years. The reasons start out easy enough, with situations that would enable us to go back to an old system if the new one doesn’t work out (I’ve never taken that option myself). Or possibly a mass influx of old server systems from enterprise consolidation projects that have performed physical-to-virtual (P2V) conversions have caused your local stash to grow.

Whatever the reason, we all have older goods in some capacity. Whether or not we’ll ever use them, that remains to be seen. This practice, while it is somewhat comforting in knowing that we have a mix of parts and equipment at the ready, may not lend itself well to being a practical solution for today’s IT staff. The best example of used gear being put to good use is provisioning a (functional) test and development environment. I keep asking myself, what am I really going to need this old server for? Or do I need this old drive array? Chances are, I’ll never get around to using these resources.

So then, what do we do with them? Should we maintain our own stash of parts? That can be an acceptable solution, until you run out or a part on the shelf doesn’t work. There are plenty of aftermarket service providers who can offer great parts and service options when warranties run out. I’ve used these services, and they are very practical and can be delivered by local resources that you can have a good business relationship with. In the interest of providing the best possible resources for an IT environment, I’d stay away from maintaining an independent pool of spare parts.

We should also take a look at what is currently running in our datacenters today. I’d venture to say that most datacenters have some element of virtualization in place, and that means that the old goods may not be the best resources. Further, virtualization naturally abstracts us from hardware, making migration to newer kits much easier.

The final nudge against holding on to old parts is their aftermarket value. You would be surprised how much money an Ultra 320 SCSI hard drive (especially the large ones) will get from resale. This is because these drives are simply not produced any more, yet they are still in use. Be sure to check my series of posts on how to delete data from hard drives with disk wiping tools. The same goes for old servers, networking, and other storage resources. They can pull in some money from the aftermarket. This can be used for future credit purchases or outright cash. Check with your provider's options on this.

How do you approach parts on the shelf? Keep them? Working diligently to remove them? Share your comments below.