Pointers on (simple?) database design

[Narj]

Hi chaps,

I've written a tool that takes network inventory from all our datacentre kit, figuring out which IP/DNS-Name is plugged into which switch and port (and against which MAC address + Vlan), but now I want to get this into a database to enable me to track changes over time (ie: stuff moving from one port to another etc). Eventually with the aim of adding a Web front end...

I've used a bit of MySQL and SQLite3 before but I haven't a clue how to start in designing a database that will allow me to track entries over time. Is there any chance of some pointers to get me started? At first I thought that IP Address would have to be a unique key but have since realised that won't work. I have no idea how to approach the time conundrum and my head hurts!

It'd be easy if I were just keeping one entry (ie: the current one) for each IP, but that's not really that useful.

Any advice would be gratefully received.

 

[Chaos]

Stanford had a free online Database course, you might still be able to watch the lectures.

http://www.db-class.org/course/auth/welcome

 

[Narj]

Stanford had a free online Database course, you might still be able to watch the lectures.

http://www.db-class.org/course/auth/welcome

Thanks, I'll have a nosey tonight.

 

[GravyMonster]

There's also a series of videos on Lynda.com about DB design. The free videos cover the theory behind DB design, so you don't need to pay for the others as they deal specifically in Access.

If you really want to mess your head you can read up about normalisation. I'd maybe save that for a bit later once you've got the basics down. Understanding normalisation is important, not least for knowing when to break the rules to get better performance.