Neil Conway <neilc@samurai.com> writes:
> Evidently the optimizer chooses a different plan in the presence of the
> WHERE clause, but that plan appears to be pretty bad. However, it is
> tough to say more without EXPLAIN output for both queries (if possible,
> get EXPLAIN ANALYZE output -- although this might take too long for the
> second query).
And the schemas for the tables involved --- without knowing what indexes
are available it's impossible to guess much.
> Also, have you run ANALYZE recently?
Indeed.
regards, tom lane