Running the hardware at the data center level will be on top — you don’t need hot-swap drives and the like, and it’s a waste of money, but of course, if you are serious about your own development work, a low-level SME server step to ensure the safety and convenience that gives you.
I myself run my own Consulting / Development (accepted answer), and I run a small network to support this. I run an HP SME server with 250 GB RAID, 4 GB RAM, Windows 2003. It runs 24 hours a day and 7 days and does everything you expect from a server in any small business. The core network connects to the Internet using a high-quality SME ADSL account.
In addition, I am running a Linux server on similar hardware connected to the Internet by another ADSL account (just a basic receivable account) on a separate telephone line. This gives me a backup in case of problems with ADSL (sometimes it happens that one goes down and the other stays), but also means that I can start a low-level Internet server to show things to clients on a dedicated and isolated line - obviously not quality to production, but good enough for testing (I have several dedicated servers co-located on other sites that run client sites, but they cannot be widely used for non-production purposes, obviously, and with built-in app atnye means I can easily configure longer)
Actually, the question of setting up your hardware requirements to support this is the same as actually deploying any hardware in the client. As a developer, you will need security and convenience, just like any small business, but equally, you need a good return on investment - you should not spend $ 1,000 on additional gold server coverage if you buy a 30-inch monitor, for example. you have a critical but understanding business partner (from the Joel coder-focus level), to whom you also have to justify your expenses, and be honest with yourself over what you need - by indulging indulgence.
EDIT in response to reading some other comments that you do not need server hardware for your personal use. Indicate that you should ask what happens when your alternative suffers a hardware failure? I believe that having a separate server makes it more likely, you can deal with a catastrophic failure faster: it can happen less often, in any case, you are more likely to get backups, and recovery will most likely be easier.
Even if you just mess with personal development, lose your job, and you have to restore complex configurations, it takes time and is a demoralized pain. You should evaluate your time at a reasonable speed of $ / h, even if it’s just fun stuff and the factor that the cost of the base server and the associated backup mode is with the prices of the equipment at a level at which they are really useless.