Re:Use SQL subqueries to count distinct 50x faster
by slowmover on Friday January 24, 2014 @05:11PM
(#7118485)
For all its faults, this is one thing I like about Oracle. No matter how many levels of nested correlated subqueries or other convoluted junk I throw at it, it (almost) always untangles the query into a nicely optimized plan. Of course, this makes it less likely I'll spend time actually thinking about how the query could be cleaned up and streamlined. Which means if I tried to move over to MSSQL or MySQL my queries would probably take forever or fail completely. Ergo, I'll stick to Oracle... ?
[ Parent
| on HN ]