Original thread is:
https://www.postgresql.org/message-id/flat/196f1e1a-5464-ed07-ab3c-0c9920564af7%40postgrespro.ru Following Yugo's advice, I have splitted this patch into two:
1. Extending auto_explain extension to generate extended statistics in case of bad selectivity estimation.
2. Taken in account extended statistics when computing join selectivity.
Now this thread will contain only patches for join selectivity estimation.
However,
IIUC, the clausesel patch uses only functional dependencies statistics for
improving join, so my question was about possibility to consider MCV in the
clausesel patch.
Sorry, do not have idea right now how to use MCV for better estimation of join selectivity.
--
Konstantin Knizhnik
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company