On Wed, Aug 31, 2011 at 4:49 PM, Andy Colson <andy@xxxxxxxxxxxxxxx> wrote:
On PG 9, after I ANALYZED the tables, it used indexes:
QUERY PLAN
---------------------------------------------------------------------------------------------------------------
Merge Join (cost=1.59..82778.35 rows=13171 width=20) (actual time=0.066..1076.616 rows=12966 loops=1)
Merge Cond: (a.id = b.id)
Join Filter: ((a."time" >= '2011-08-15 00:00:00'::timestamp without time zone) OR (b."time" >= '2011-08-01 0
-> Index Scan using a_pkey on a (cost=0.00..31389.36 rows=1000000 width=12) (actual time=0.007..204.856 ro
-> Index Scan using b_pkey on b (cost=0.00..31389.36 rows=1000000 width=12) (actual time=0.006..224.189 ro
ANALYZE is the magic.
You are, of course, right, but it doesn't use the TIME index.
Hmmm, may be my example isn't large enough to produce the issue.
--
Eld på åren og sol på eng gjer mannen fegen og fjåg. [Jøtul]
<demo> 2011 Tore Halvorsen || +052 0553034554
