On Fri, 26 Dec 2008 19:13:48 -0500 Tom Lane <tgl@xxxxxxxxxxxxx> wrote: > The subselect syntax certainly seems like the one most likely to > work across different SQL implementations. WITH is a pretty subselects actually works on mysql too but on a 1M table with about 300K unique columns it performs more than 4 times slower than select (distinct a,b) from table 18sec vs. 4sec Times were similar for innodb and myisam. Postgresql needs 17sec with subselect. I didn't try to see how both db could perform with indexes. mysql performance is impressive. I thought that most of the time would be spent on "distinct" where postgresql shouldn't suffer from its "count" implementation. But well still 300K rows to count on 1M aren't few. -- Ivan Sergio Borgonovo http://www.webthatworks.it -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
