I have a large, single table database that keeps track of daily rainfalls from over 7000 locations around the country. This means that for every day, the year,month,day and station name are repeated. The database gets bigger each day and performance is being badly affected more and more (not surprising). I do a lot of SQL querying with the data. I know I could archive some of the older records but really that solves nothing in the long run. The table presently has over 3.5 million records and takes nearly half a minute to open. Compressing the database makes no difference.
The problem relates to efficient design. Although I've read up a fair bit on database design I really don't know how to sort it out. I'm really scarcely a novice at this.
Would there be any advantage in keeping seperate tables for unique days, months, years and station names, as well as the original table of data (downloaded daily so I have no control over its initial format)? If not, could someone please suggest an alternative especially so that tables and queries would run faster?
Thank you
The problem relates to efficient design. Although I've read up a fair bit on database design I really don't know how to sort it out. I'm really scarcely a novice at this.
Would there be any advantage in keeping seperate tables for unique days, months, years and station names, as well as the original table of data (downloaded daily so I have no control over its initial format)? If not, could someone please suggest an alternative especially so that tables and queries would run faster?
Thank you