Home/ Questions/Q 8311769
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-08T19:55:13+00:00

Our application has many configuration directives. Things like are multiple languages enabled? or which

  • 0

Our application has many configuration directives. Things like “are multiple languages enabled?” or “which CSS template should be used?” or “how many items should be shown per page?”.

We need to store the configuration directives in DB. Currently we have table for each “area” of configuration and the table has only one row containing the values in columns. This approach allows us to put constraints on the values – let it be ENUM, INT, VARCHAR, etc. But it’s huge PITA when we need to add new config directive — we need to update and redeploy DB schema. It also pollutes the table space with many setup_something tables.

My idea was to replace the tables with MySQL based key-value store (we use shared hosting with little configuration options, so we cannot use NoSQL). Some of the values MAY depend on locale or other variables, but I think that it can be solved by something like fallback (find key begining with “lang:fr” and if not found try one begining with “lang:default” (which will be always ready).

The constraints would be checked only in PHP using concrete setup classes, but there will be no DB checks.

  1. Are there any ready-made solutions for this (prefferably working well with Zend Framework).
  2. What are the pitfals of this solution (except the obvious – that I can’t fetch the whole “row” at once or many rows at once – which is not a problem)
  3. [Have you used|Would you use] similar solution or something completely different?
Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    You could simply have one MySQL table with all your config settings. Perhaps something like this:

    config_id (INT autoincrement)
config_group (VARCHAR) this gives you something to group your configuration setting with
config_key (VARCHAR) this would be the key name by which you would reference the value
data_type (VARCHAR) you can use this to tell your application how you want to treat this data within the application (i.e. treat it as integer, float, string, etc.)
value (VARCHAR) the value

    Of course you would then need to take the string value and manipulates it in your code to treat it like an integer, float, string, etc.

    Maybe note the cleanest way to do things, but have seen this approach taken a number of times.

    • 0