|
Well.... having used BOTH in extreme situations (millions of records with complex joins), I can add a little to this.
MySQL is fast for straight queries. Throw a join into the mix and it's slower. Throw a multi-dimensional join in, and you're dead meat. it could easily peg you down to 4 minutes. I could run a query on MySQL for 300,000 records that would take over an hour (a row-inequality self-hash, for example)
Oracle, with sub-select capabilities and stored procedures for compiled, optimized queries, and clustered indexing for faster joins on indexes is an incredible difference (for someone who knows how to use it). Its downside is that it's very difficult to learn. If you do learn it, though, you can VERY feasibly turn a 4 minute MySQL query into a 2 second procedure.
Last edited by smoseley; June 27 '03 at 10:14 AM.
|