Ask HN: Learning to tune a database

[catch404]

I've been wanting to expand my knowledge around databases and sql/tuning. I've worked mostly with Oracle and would like to learn Postgres.

My problem is: With a programming language, I can do the 'Hello world' app and progress from there. With a database I need a different problem to solve or a different dataset to work with. I'm not going to learn much about tuning if I release a little used webapp :)

I bet there are plently of datasets around that I could load and mash up but I'm not too such on where to start.

Any suggestions?

[wooster]

Sure. Make an app that stores a very large amount of data. A good example would be spidering web data (unstructured), parsing some structure from it, and storing it in a database. Then, do some analysis on it, expose it to the web, etc.

The important thing is either to:

a) Have a lot of data.

or

b) Have a lot of traffic.

The first is a lot easier to get than the second.

[trapper → wooster]

And both are simple to simulate :)

[codeodor]

I think you want more than a lot of data.

That's a good start, but you also want it to be structured in a way that requires some thought (or just in 3+ normal form). Have someone ask for a report out of it (or come up with as many reports as you can).

I wish I had a suggestion for such a data set. Microsoft's famous case is Northwind, but I don't know where you might find it (without having searched just now).