Thanks to take a look!
On 11/25/24 23:45, Heikki Linnakangas wrote:
On 24/09/2024 08:08, Andrei Lepikhov wrote:
+ * proves the var is unique for this query. However, we'd better
still
+ * believe the null-fraction statistic.
*/
if (vardata->isunique)
stadistinct =
On 24/09/2024 08:08, Andrei Lepikhov wrote:
On 19/9/2024 09:55, Andrei Lepikhov wrote:
This wrong prediction makes things much worse if the query has more
upper query blocks.
His question was: Why not consider the grouping column unique in the
upper query block? It could improve estimations.
Af
22b572903b8c1a2459b051f91e0902a2fc243648 Mon Sep 17 00:00:00 2001
From: "Andrei V. Lepikhov"
Date: Wed, 18 Sep 2024 19:25:02 +0200
Subject: [PATCH v2] Improve statistics estimation considering GROUP-BY as a
'uniqueiser'.
This commit follows the idea of the 4767bc8ff2. GROUP-BY make
V. Lepikhov"
Date: Wed, 18 Sep 2024 19:25:02 +0200
Subject: [PATCH] Improve statistics estimation considering GROUP-BY as a
'uniqueiser'.
This commit follows the idea of the 4767bc8ff2. GROUP-BY makes the grouped
column 'highly likely unique'. We can employ this fact in
