I have some settings files, that will differ in naming of the different settings.
setting1 : true
setting2 : false
setting3 : false
settingsss1 : true
settingsss2 : true
settingsss3 : true
the naming is user generated so cant really make columns for them.
i though doing something like making a table with
Id - Key - Value
This would then give 50 entries for a settings file with 50 settings.
Would it be better to just save these as JSON string in a field in the DB as MediumText?
i expect to have around 100 new saves of these every day. giving 100 x 50 x 365 = 1.800.000 entries a year..
The thing is I need to be able to look thru these in a decent way, to compare their settings.
should I go with JSON string or entry pr. parameter. or even a 3rd option?
>Solution :
For simple, flat key/value pairs with simple values, use a normal table. If the values are just true/false use boolean for the value column to save a little space.
If the values can be a mix of types (dates, numbers, strings, lists, nested key/value pairs) use the JSON type. It can be indexed and searched.
Millions of entries a year is not a lot if they are properly indexed. Handling large amounts of data is what SQL databases are good at.