Re: HASH JOIN - how does it work?

My reply is based on the tuning I've been doing lately, so it's not a technical description. A hash join is done instead of a nested loop by doing a full scan on one of the tables (the optimizer decides). This sounds ugly, but the hashing algorithm usually does it better than nested loop. So far, I've never been able to beat the optimizer's choice of hash over a nest loop: every time I've forced a nested loop (/*+ USE_NL */), it was worse than the hash. Received on Wed Apr 01 1998 - 00:00:00 CST