Database Reference
In-Depth Information
F
Vacuum and index maintenance: To reduce bloat, including collecting optimizer
statistics through ANALYZE.
F
Consider VACUUM again, with the need to manage the less frequent freezing
process. This is listed as a separate task so that you don't ignore this and have it bite
you later.
F
Server log file analysis: How many times has the server restarted? Are you sure you
know about each incident?
F
Security and intrusion detection: Has your database already been hacked? What did
they do?
F
Understanding usage patterns: If you don't know much about what your database is
used for then I'll wager it is not very well tuned or maintained.
F
Long term performance analysis: It's a common occurrence for me to get asked to
come and tune a system which is slow. Often what happens is that a database server
can get slower over a very long period. Nobody ever noticed any particular day when
it got slow, it just got slower over time. Keeping records of response times over time
can help confirm whether or not everything is as good now as it was months or years
previously. This activity is where you might reconsider current index choices.
Many of these activities are mentioned in this chapter or throughout the rest of the cookbook.
Some are not because they aren't so much technical tasks but more just planning and
understanding of your environment.
You might also find time to consider the following things:
F
Data quality
:
Are the contents of the database accurate and meaningful? Could the
data be enhanced?
F
Business intelligence: Is the data being used for everything that can bring value to
the organization?
