Re: Very Large Non-Dynamic Table for Lookups

[Eidetic Hans]

You're right, not all SSNs have been allocated. I actually only have around 10,000,000 but the integrity can't be vouched for. The reason it may go up is because I can generate SSNs and randomly assign UIDs for future projects. What I'm mainly interested in is using MySQL as a basic retrieval system for large data sets (not a full transactional or even relational system).

The company or department IDs didn't come with the data. That's a good question, though. Maybe they were stripped before being given to me.

100 disk hits a second sounds costly. Since the table won't increase anymore would there be a way to minimize this? If I sort the SSNs, would that make a difference?

I'll increase the innodb_buffer_pool_size to 70%. I'm running XP with 2GB of memory (not ideal) but this would be a dedicated action.