On the last month I’ve done basically the impossible: I have a Debian server on a Intel Celeron 2.5Ghz / 512 MB RAM / >40GB IDE Hard Drive with MySql running smoothly. I managed to connect using MySql Workbench and then I realized that I didn’t stop to think about the database model.
My current database is an Access 97 Database with 2 gigantic tables:
Tbl_Swift – 13 fields and one of them is a ‘memo’ field with a full page of information.
Tbl_Contr – 20 fields where FOUR of them are ‘memo’ fields with pages of information.
It’s not like the database is heavy or slow on Access, but I wanted to make it available to most users… then I realized that I should optimize my database, but here’s the problem:
WHY?
Will it make that much of a difference? I’ll have less than 5 users connected to this database and NONE of them will have ‘write’ privilege, they’ll just run some standard queries. The database itself is rather small, it’s under 600MB and ~90K records.
So, should I really stop to think about making it more ‘optimized’?
“When I say OPTMIZE I mean when people say I should have a lot of tables with little information”
What you are talking about is normalization, and recently there was a thread about normalization vs. performance here: Denormalization: How much is too much?
And yes, I believe that you should think about normalization before that DB gets too big.