On 3/27/25 01:58, David Rowley wrote:
> I suspect the fix for this might be a bit invasive to backpatch. Maybe
> it's something we can give a bit more clear thought to after the
> freeze is over.
One more issue I think may be addressed (or just considered) here is the
following:
CREATE TABLE parted (a int, b int) PARTITION BY RANGE (a, b);
CREATE TABLE part1 PARTITION OF parted
FOR VALUES FROM (1, 1) TO (1, 10);
CREATE TABLE part2 PARTITION OF parted
FOR VALUES FROM (2, 1) TO (2, 10);
INSERT INTO parted (VALUES (1, 2));
INSERT INTO parted VALUES (2, 2);
EXPLAIN (COSTS OFF)
SELECT * FROM parted WHERE a > 1 AND b < 1;
EXPLAIN (COSTS OFF)
SELECT * FROM parted WHERE a > 1 AND b > 10;
/*
Seq Scan on part2 parted
Filter: ((a > 1) AND (b < 1))
Seq Scan on part2 parted
Filter: ((a > 1) AND (b > 10))
*/
I think partition part2 could be pruned like in the following example:
EXPLAIN (COSTS OFF)
SELECT * FROM parted WHERE a > 2 AND b > 10;
/*
Result
One-Time Filter: false
*/
--
regards, Andrei Lepikhov
