Thanks to everyone who replied to my email about a
subselect query that was rather slow (1-3mins). The query is:
SELECT * FROM items WHERE package IN (SELECT package FROM items WHERE ...user search
expression... GROUP BY package)
The fastest query I could come up with that is equivalent
to the one above is
SELECT DISTINCT i1.* FROM items i1, items i2 WHERE i1.package = i2.package AND ...user search
expression...
which finishes in about 3s. EXPLAIN gives the following info on
the query (if "search expression" is a regexp on non-indexed
column i1.performer):
Unique (cost=1334.61 rows=2 width=151) -> Sort (cost=1334.61 rows=2 width=151) -> Nested Loop
(cost=1334.61rows=2 width=151) -> Seq Scan on items i2 (cost=1332.56 rows=1 width=12) ->
IndexScan using items_packages on items i1 (cost=2.05 rows=12805 width=139)
But it is still faster if I run N+1 queries - first one SELECT
query to list packages, then another SELECT for each row in
the result.
Oskar Liljeblad (osk@hem.passagen.se)