The answer to this question depends on how fast you want your data back and how much time and effort you are prepared to spend. If your data is both mission and time critical, then full, frequent backups possibly with mirrored systems for immediate restore or failover may be the only solution. Financial trading organisations, large volume e-commerce sites and hospital emergency wards are examples. Other users who do not want to or cannot go down this route will be faced with more basic options.
- Isolation. If a storage device is still working, the system to which it is attached will likely continue to try to store data on it. This may in turn overwrite the data you want to recover. Switching off the system where possible, and/or removing the storage unit are therefore important first steps.
- Repair. A storage device that has ceased to work properly (includes strange clicking and beeping) may be repairable, at least to the level of making a stable image of the data it holds. This device clone can then be worked on safely to restore data as much as is possible.
- Recovery. There may be a reliable data recovery application available for your computing installation that also avoids any additional data loss. For MS Windows systems for example, it is possible to boot up from a CD-ROM containing a recovery utility to transfer files to another medium without interfering with the original disk’s contents. The ‘Chkdsk’ function within MS Windows itself fixes logical faults in disk storage to help recover data as well.
Should users hang on to obsolete systems for which data recovery can be an arduous and uncertain process? Both support costs and operating problems tend to increase as IT systems age, and the situation may become not only risky but also uneconomical. However, some organisations find themselves hostage to a system developed in-house or on exotic hardware with complex business logic built in. Migrating to better solutions then takes time, during which the options above may be the only practical way of managing data storage failure.