It will be inevitable that a tornado is going to hit you when you live in a tornado state. So it makes no sense to not worry about tornadoes when it is more than likely going to happen. Same applies to a hard drive failure on dedicated servers (and other miscellaneous hardware failures.) Since there is a good chance this can happen to an older hard drive or a brand new one, (take a look at bad luck here it is vital to make sure the hard drive can keep your data safe as much as possible and that you have good website backups in place at all times and that you know in advance that a failure might occur.
1) The Windows Operating System can sometimes indicate a pending hard drive failure on a server. Tech support should suggest reinstalling the OS on a new drive. Having backups can make this process easier and a quick restore should be possible. Simply utilizing more expensive drives that have better resiliency than cheaper ones and investing in a RAID system can also help. A best-case scenario is data backed up to another off server location. This is ideal since this can also be password protected against intrusions and hackers trying to obtain access to sensitive data.
Another good tool to use is SMART and there is more information on this here.
2) Performing periodic disk defragmentation can be a slow but easy way to avoid or even predict a hard drive failure. Additionally, it can speed up server performance a good deal. Fragmented disks end up having to read much more as it “hunts” or seeks for the requested data. Contiguously stored data speeds up read and write performance and can avoid early disk failure. Scheduled disk defragmentation on a regular basis. It should run easily on off-peak hours of the day at least a few times a year depending on the amount of usage the hard drive experiences. Some dedicated server administrators will want to run disk defragmentation more often than others.
Prevention instead of imminent hard drive failure
A dedicated server admin always wants to try to prevent or limit the possibility of a hard drive failure and not have to deal with it when it does happen without a plan of action or without any backups. It is always best to create an environment that limits these problems from happening. As hard drives become larger and larger and prices come down, restoring a large drive will indeed take much longer. This will cause longer periods of downtime which is something to consider.
Plan a hard drive test at ideal times. Not when hard drive failure is suspected and a test needs to be run at an inopportune time.
A newer type of data protection is based on Error Correction Codes, ECC. This is all performed “on the fly” and no user intervention is required. However, manufacturers do take into account how much correction is needed. This can have an impact on reading speeds, processing power and physical storage space possible on the drive. For more information on ECC please go here.