Re: big database +1 million rows/days

[Rick James]

In order to answer your question, we need to know what you will be doing with the data. Do you have the SELECTs that will be fetching it?

Normalization will shrink the 'fact' table. Shrinking it will save space, hence I/O, hence speed. Normalization may require some code; are you using Perl/Java/etc?

int(30) does not make sense. How big will the numbers really be? Probably INT UNSIGNED will work for you; it has a max of 4 billion.

What column or combination of columns is unique?

KEY `value` (`value`),
KEY `value_2` (`value`)
Those are redundant; DROP one of them.

> 1 million rows/days
Which table are you referring to?

> Object: parameter1 = value1 , value2, value3,....;
Too vague; please show use the SQL and/or sample of the data.

Don't use FOREIGN KEYS, that slows things down.