Category: Articles

Denormalised Design: Prime Numbers and SQL

Many may think that prime numbers are for the mathematicians to play with, and bit of an obscure combination to use with SQL. However, they can be very useful for implementing a compressed data solution providing some additional set theory operations difficult to achieve in SQL.

Database Design: Block Size Design

A database designer has a number of factors to consider when determining a tablespace strategy for a database. A big influence on this is the block size of a tablespace, and if not the default size, how they will work with the various caches.

Data Generation: Techniques in SQL

Software applications need to be tested and this may mean being dependent on a lot more data than can be entered through natural means. The quality and authenticity of generated data relies on the sophistication and effort put into the routines. This article describes some techniques to ease the coding.

Database Design: Missing or Bad Statistics

A common cause of poor SQL performance in an application is missing/stale CBO statistics on a table. The typical aim is to have statistics gathered on all tables belonging to an application; however there are situations where statistics are not possible or are inappropriate. This article covers the scenarios where this occurs and proposes methods to ensure that they are applied effectively.

Database Design: Block Contention

A common bottleneck on busy OLTP applications is contention on the buffer cache. If too many sessions attempt to read the same blocks from the buffer cache, then they can queue up waiting for low level latches to be released before they can read the blocks from the cache.