Databases Reference
In-Depth Information
T
HIS
SECTION
DESCRIBES
ANOTHER
LEVEL
OF
PERFORMANCE
TUNING
FOR
.
DATABASES
AND
APPLICATIONS
TO
IMPROVE
APPLICATION
RESPONSE
TIME
Overall application performance is the result of the sum of the individual
performances of each component involved in a transaction, plus some over-
heads. Performance tuning is an iterative process that requires extensive
testing to locate bottlenecks and other exposures anywhere in a solution. A
well-respected performance tuning approach involves tuning components
that give the biggest impact on the bottom line, and then continuing the pro-
cess until application performance is within an acceptable range.
Performance tuning in Web-based applications requires traditional
techniques, such as database tuning, indexing, locking, and efficient algo-
rithms. However, the Web-based components require additional approaches,
techniques, and tools to optimize application performance. This section
provides a cross-section of chapters that examine approaches for improv-
ing various components in an overall architecture to maximize application
performance.
Chapter 44, “Improving User Experience Through Improved Web Design
and Database Performance,” examines considerations to improve overall
application performance and Web site design. This chapter offers sugges-
tions for improving the overall user experience by improving application
performance and the appeal of the user interface layout and user interaction.
Chapter 45, “Web-based Testing and Capacity Planning,” examines some
of the key differences in testing a Web site compared to other technical
architectures. The chapter also examines some of the specific capacity-
planning issues facing Web design. Transaction volumes against Web sites
are unpredictable, as the potential user base consists of anyone with a
computer and a password. Because many Web sites allow users to get an
account with relative ease, the number and complexity of operations
require a more sophisticated approach and testing harness than do other
non-Web architectures.
Chapter 46, “The Advanced Art of Indexing,” describes the develop-
ments and challenges of indexing database tables to improve the perfor-
mance of operations against the tables. Some operations, such as joins,
updates, and subqueries, can be greatly improved by traversing indexes
during parts of the transaction execution. Optimizing index utilization
requires significant trial and error, and a highly iterative approach. Various
indexing techniques such as B-Trees, bitmaps, and inverted files are exam-
ined in this chapter.
Chapter 47, “Parallel Databases,” examines opportunities to improve
application perfomance by leveraging multiprocessing and multithreading
via parallel architecture to distribute workload over a wider base of tools.
Search WWH ::
Custom Search