What'a good way to store large time/value datasets?

Question

I'm working on an application that stores a lot of quite large time/value datasets (chart data, basically values taken from a sensor every day, hour or 15 minutes for a year+). Currently we're storing them in 2 MySQL tables: a datasets table that stores the info (ID, name, etc) for a dataset, and a table containing (dataset ID, timestamp, value) triplets. This second table is already well over a million rows, and the amount of data to be stored is expected to become many times larger.

The common operations such as retrieving all points for a particular dataset in a range are running quickly enough, but some other more complex operations can be painful.

Is this the best way to organize the data? Is a relational database even particularly suited to this sort of thing? Or do I just need to learn to define better indexes and optimize the queries?

Answer 1

A relational database is definitely what you need for this kind of large structured dataset. If individual queries are causing problems, it's worth profiling each one to find out if different indexes are required or whatever.