Published on

January 16, 2010

Avoiding Time Bomb Coding in SQL Server

As a SQL Server developer or administrator, you may encounter performance problems in database applications. Sometimes, these issues can be easily diagnosed and resolved, but in other cases, the causes of poor performance are more subtle. One common pattern that can lead to performance degradation is known as “Time Bomb Coding”. In this article, we will explore this pattern and discuss some examples of how it can be avoided.

The Time Bomb Coding Pattern

The Time Bomb Coding pattern typically exhibits the following properties:

  • The system is well established and has not undergone significant alterations for some time.
  • Statistics and indexes are up-to-date.
  • Execution plans appear reasonable at first glance.
  • The database design seems simple and clean.
  • Traffic has not changed significantly over time.
  • Despite these factors, the system’s performance has been steadily degrading.

If you have experienced this pattern, you may be a victim of Time Bomb Coding. Let’s explore a few examples to better understand this phenomenon.

Example: Inefficient Count Queries

One common example of Time Bomb Coding is the use of inefficient count queries. Consider the following code snippet:

IF (SELECT COUNT(*) FROM ... WHERE ...) > 0
BEGIN
    -- Perform some actions
END

While this code may seem simple and clear, it can become a performance bottleneck when executed against a multi-million row table. The accurate count of records is not relevant; only the existence of records satisfying the WHERE clause matters. By changing the condition to IF EXISTS(SELECT 1 FROM ... WHERE ...), the application’s performance can be significantly improved.

Consider Timeliness when Providing Aggregates

Another cause of Time Bomb Coding is the failure to consider when it is necessary to provide aggregate values such as counts, sums, or averages. For example, if a website displays the number of reads for each article, calculating this aggregate value on the fly can become increasingly expensive as the number of reads grows over time. By evaluating the events that cause these aggregates to change, we can determine when the calculation of aggregates should take place. This can be achieved by maintaining a summary table with appropriate triggers or batch processes.

Avoiding the One True Lookup Table Problem

The “One True Lookup Table” problem is another common issue that can lead to Time Bomb Coding. This problem arises when a single table is used to store all lookup values for different entities. While this design may seem elegant at first, it can quickly become unmanageable and hinder data referential integrity. Instead, consider separate lookup tables for each entity or use views to present a unified view of similarly structured entities.

Conclusion

Time Bomb Coding can have a significant impact on the performance and maintainability of SQL Server applications. By carefully considering factors such as accessibility, accuracy, completeness, consistency, integrity, relevance, and timeliness, developers and administrators can avoid falling into this pattern. By following best practices and avoiding common pitfalls, SQL Server applications can maintain optimal performance and scalability.

Have you encountered Time Bomb Coding in your SQL Server projects? Share your experiences and strategies for avoiding this pattern in the comments below!

Click to rate this post!
[Total: 0 Average: 0]

Let's work together

Send us a message or book free introductory meeting with us using button below.