> Ran analyze on both. Running the same query I'm getting different plans, one x10 slower.
theory:
the "statistics target" is too low ?
THEN
different random sample --> different statistics ---> different plan,.
"For large tables, ANALYZE takes a random sample of the table contents, rather than examining every row. "
IMHO: Try to increase the statistics target ( "default_statistics_target" , .. )
regards,
Imre
I have two identical databases running in the same instance of Postgresql. Ran
analyze on both. Running the same query I'm getting different plans, one x10 slower.
Although I have solved my problem by re-writing the query, I want to understand why
this is happening. If the configuration, Postgresql version, schema and data are the
same, what other factors is the planner considering?
--
Kostas Papadopoulos
KE MethodosIT