Creating a recovery manual

As part of your system recovery plan, and to maximize the reliability and availability of your servers, you should create a recovery manual for each computer. The recovery manual should contain all the information necessary to recover the computer.

Provide both a printed copy and an electronic copy of the recovery manual. Keep a printed version near the server. If possible, physically attach the printed copy to the server to ensure that the copy does not get lost. Make the electronic copy available at the operations center so that it is readily accessible to all operators and technicians (including remote operators). In addition, consider making printed copies available at remote locations in case network servers become unavailable.

The recovery manual should contain the following:

• All configuration information for the hardware and software.

  This should include all information required to rebuild the computer, such as the Internet Protocol (IP) address, computer name, interrupts and input/output (I/O) settings for any network adapters, disk controller configuration settings, the amount of memory, and the number and types of cards and the slots they occupy.

• A set of media (CDs or other media), including the drivers, applications, and operating system needed to reload the computer.

  As an alternative, you can provide links to network servers that contain copies of necessary drivers and applications. In addition, however, at least one local source of drivers and other software should be made available in case the network servers become unavailable.

• As appropriate, provide copies of all hardware documentation (manuals for all hardware components).

  If the hardware manuals include information about more than one component model or component type, highlight or note the component model that is installed on your computers. It can also be useful to add installation notes in the margins of the documentation.

• Provide procedures for diagnosing, restoring, and repairing the computer.

  Include instructions for diagnosis and for choosing the best type of response if a problem arises. These instructions might include lists of log files to check, including operating-system log files, application log files, and other logs.

The printed copy should be updated whenever a change is made. In addition, reviewing and updating printed recovery manuals should be part of a routine monthly survey of the computers.