Yeah, I asked it. I have a ridiculous amount of data I would like to "join" together like in SQL and I'm a poor trash retail trader who learned about the stock market last week. I've seen a few threads on people doing various things for tick databases, but let me be real specific: 1) Want the ease of SQL-like queries 2) Need about 500-700 million rows, and about 50-80GB of data per "instance" (instance = research project) 3) Query time for a slice of time (at most a few months) for any given ticker should be a few seconds at most 4) Needs to be usable for a guy who knows NOTHING about databases, other than the online tutorials about SQL 5) Works on a non-server computer that's about 3-4 years old, 4GB of ram tops 6) Learning curve can't be so high that I spend a month reading manuals I'm homogenizing tick data into time intervals, so it doesn't have to be tick for tick. I just need something that works with a reasonable amount of speed. Simple and "fast enough" is all I care about. The system does not have to be available during the day and is for research. I've read the threads on ET already, but am not competent enough to understand all of the lingo and implications of design decisions, so please respond as if you are addressing a total beginner. What is the best way to get as close as possible to what I want, using the cheapest, most-free stuff, as quickly as possible? Thank you.