Over the years, many readers have responded with questions to a previous article I wrote on how DBAs can move from a reactive to a proactive environment by developing and implementing best practices ( see “Database Administration Best Practices,” z/Journal, October/November 2003):
• Why do you fail at creating best practices?
• How does your organization compare with others?
• Just what are best practices?
Best practices emerged during the continuous improvement and total quality management movements of the ’50s. In manufacturing, they were created as responses to requirements or derived from business strategies. A best practice will commonly consist of three characteristics: documentation, process, and process discipline (standards).
Documentation—that is, other people’s documentation—should be familiar. We rarely seem to create our own! Documentation for a best practice must exist in a readable format and be stored in a centralized location accessible by everyone, not only those most likely to read or use it. There must be a standard method of finding best practice documentation. It must be regularly reviewed by the owning team for timeliness and accuracy. Finally, best practice documentation must have a quality measure.
Quality measures for documentation go beyond good grammar. Teams must decide what quality measures are important to them. Some of these might be:
• Applicability: Does this documentation apply to many general conditions or only a few specific cases?
• Verifiability: Can the documentation be cross-referenced to other documentation or publications such as product manuals? Are there means of verifying that the documentation is correct, current, and complete?