Companies operate under a significant amount of pressure. Logistics nowadays are implemented in a way which inherently means any possible downtime and software/hardware failure would bear disastrous consequences for the business.
The ability to recover quickly from any type of component failure is critical, particularly in the case of servers or databases; this prevents or significantly reduces downtime, effectively preventing losses as well.
Depending on the warranty provided with the equipment, in a complex business-operating system, a component failure can mean waiting several days or even weeks for part replacement. How quickly the component will be replaced will depend on a number of factors, such as service level agreement and the ability of the service provider to quickly replace the part. Tier one manufacturers offer a range of warranty options, of course, for your consideration – still, the trick here is in the details. The details in a warranty are usually typed in fine print, meaning that the response time of, say, two hours, is exactly that – the provider will respond to your distress call within two hours. In most cases, however, the part replacement will take significantly longer.
Depending on the location of your business, the shipping and handling could cost an exorbitant amount. Another downside is that there is no recourse if the manufacturer doesn’t fix the broken/damaged part. Needless to say, people are rarely aware of this, and it isn’t till a disaster occurs that it becomes painfully apparent just how important it is to react quickly. Even in case you pay a hefty fee to make your warranty as efficient as possible (with regard to the options the manufacturer is offering), it is still likely you will wait for days before the replacement for the broken part arrives.
An alternative to any potential ordeal is to implement a high availability infrastructure, which effectively means having two of everything on-site. We will assume your company implements good practice with hard disk drive redundancy with a mirrored operating system on one set of disks, and a RAID 5 configruation (Redundant Array of Inexpensive (Independent) Disks), with striped blocks and distributed parity on data drives.
One way to implement high availability is to purchase a second server chassis, processor board and memory, and have everything racked side by side with the production server.
This option is most likely cheaper than the 4-hour maintenance agreement from your manufacturer and will be faster to recover from in the event of server failure. Should components in the production server fail, it’s fairly simple to switch the hard drives to the redundant chassis. No additional software licensing is required, which is particularly important, since, beside money, this saves your business a lot of legal and other trouble.