All departments should develop and maintain procedures that will ensure an appropriate level of customer service in the ......... of unexpected business systems failure.

Traditionally, disaster recovery planning focused on the physical aspects of computer-based systems, particularly the CPU and closely coupled ........., e.g. disc banks.

Although this approach remains valid, the ......... of disaster recovery has considerably widened in recent years.

This change of emphasis has resulted in the need for a pragmatic service level delivery rather than plans, ......... documented and often untested.

Departments are now so reliant upon computer-based systems for normal daily operations that small systems failures can cause ......... level of disruption to customers and staff.

A balance must be ......... between a 'business as usual' approach, with its associated high cost, and recourse to a 'survival' mode of operation with the objective of 'business as usual' within a reasonable period.

Due to the wide variety of business applications across departments, it is not possible to provide ......... guidelines for the management of unexpected failure and disruption of computer-based business systems; there are, however, some general guidelines that are globally applicable.

All ......... plans must be tested; critical components of the plan should be tested once a year, and test plans must ensure that the testing itself does not cause a 'disaster'.

Plan for key personnel not to be present at the test — this aspect is often overlooked, with all key personnel ......... days beforehand for the test.

......... the 'walk-through' method for testing the plan prior to putting it into operation; do not declare the testing completed until the whole production system has been re-instated.