Microsoft Word Hardware Reconfiguration Methodology V final2

Alternate methods of restoration

It is important to ensure, as with backups, that the restoration plan examines and outlines various alternative methods for restoring data back to the required system(s). As with backups, it can happen that backup devices or network paths become unavailable and another method for backing up the system(s) must be found. It is the same with data restoration; unforeseen and

uncontrollable circumstances can cause data restorations to become very complex and cumbersome, and when being performed against mission critical systems, time is essential. For example, if certain disks or filesystems cannot be brought offline on the system marked for restoration, then data can be restored to an alternative directory (on the system or on the network somewhere else and shared) and copied over at a more appropriate time. Alternatively, if network restorations are performed but the network becomes inoperable then it is important that the backup device and media can be easily connected and reconfigured for the system that is to have its data restored.

6. Procedure testing

It is very important to test all restoration procedures before ever requiring them. Testing should be done in a laboratory or other controlled environment where different tests and scenarios can be experimented on without affecting the operational network. Procedures should be tested to determine their suitability and adaptability to changing contextual situations in order to understand how to improve upon them (when and where necessary) so that they can be made to accommodate multiple uses. In addition, by testing them it becomes possible to understand how they will behave due to varying circumstances caused by, for example, a lack of system or human resources, device failure, and network unavailability. However, not all procedures are equally useful nor are they equally applicable in all circumstances. For instance, backed up system data that is stored on tape that is no longer accessible due to a failed local SCSI controller will have to be restored over the network. Another example would be a system’s main operating system and boot disk that has failed. It will have to be replaced with a spare disk and its data reloaded from backup; this will require booting from a rescue or Live CD. In the latter example, procedures should already be in place for how to use alternative boot devices and how to restore data using these devices. Regardless of the varying circumstances, it is important that the backup media used can be recovered using the various procedures and these procedures should be found within the restoration plan. The most difficult issue about a restoration is adapting to varying problems and circumstances when time is running out for a mission critical system; a set of procedures will likely help to better adjust to such circumstances.

7.