How a Simple Diagnostic Box Can Reveal and Fix Slow Database Queries

I will share a series of useful and easy-to-implement tips accumulated during my development career; this article is the first in the series.

Many factors can cause a website to run slowly, but the most common is excessive database query time. Database queries often play a crucial role during page rendering, and unnecessary or unoptimized queries can degrade performance.

Typical problems include:

Queries whose results are never used.

Queries that do not use indexes.

Multiple queries where a single query would suffice.

Slow and complex queries.

However, a simple mechanism can indirectly prevent and fix these issues:

Use a “diagnostic box” on every page to show the number of database queries and the total time they consume.

Below is an example of a diagnostic block from one of my own sites:

Request Details: DB – Queries: 4, Time: 5.66 ms

I have added similar diagnostic boxes to all development projects and production environments (when logged in with my account). For large sites you may enable the box only when accessing from the office or VPN. The box can also be extended to show the actual SQL statements and other timing details.

Benefits

I can instantly see whether the database is causing slow page loads, which is especially valuable when debugging in production, as query performance can differ dramatically between environments.

When adding new features, I can visually assess whether the additional queries are lightweight, helping balance functionality against data access cost and reminding me to optimize or cache expensive queries.

If a project uses an ORM, the box reveals whether the generated queries match my expectations.

When removing a feature or adding caching, I can confirm that the query count has decreased as intended.

Conclusion

These ideas are not original to me; many large sites already implement similar diagnostics. If you haven’t adopted them yet, I strongly recommend taking the time to implement a diagnostic box. While it won’t make the site itself faster, it encourages better development habits and can have a huge long‑term impact on performance.

If you’re interested, you can view the SQLAlchemy query‑count code I use on the site mentioned above.