I'm using 8.2 and using order by & limit is still faster than MAX() even though MAX() now seems to rewrite to an almost identical plan internally. Count(*) still seems to use a full table scan rather than an index scan. Using one of our tables, MySQL/Oracle/MS-SQL all return instantly while PG takes longer ther 700ms. Luckily we can design around this issue. -----Original Message----- From: pgsql-performance-owner@xxxxxxxxxxxxxx [mailto:pgsql-performance-owner@xxxxxxxxxxxxxx] On Behalf Of Craig A. James Sent: Sunday, January 07, 2007 5:57 PM To: Guy Rouillier; PostgreSQL Performance Subject: Re: [PERFORM] High update activity, PostgreSQL vs BigDBMS Craig A. James wrote: > The "idiom" to replace count() was > "select col from tbl order by col desc limit 1". It worked miracles for > my app. Sorry, I meant to write, "the idiom to replace MAX()", not count()... MAX() was the function that was killing me, 'tho count() also gave me problems. Craig ---------------------------(end of broadcast)--------------------------- TIP 6: explain analyze is your friend
