Danil Anisimow writes:
> In some cases, the planner underestimates FULL JOIN.
Perhaps.
> Are these simple changes enough to improve this situation?
This looks like an entirely ad-hoc change, one that could make as
many cases worse as it makes better. If you want to argue for
merging it, you at
Hi.
In some cases, the planner underestimates FULL JOIN.
Example:
postgres=# CREATE TABLE t AS SELECT x AS a, null AS b FROM
generate_series(1, 10) x;
postgres=# ANALYZE;
postgres=# EXPLAIN (ANALYZE, TIMING OFF) SELECT * FROM t t1 FULL JOIN t t2
ON t1.b = t2.b;
