Hi Wei Shan,
Thank you for your response.
Query B was run after initializing the DB ex-novo doing VACUUM ANALYZE before and after creating and clustering indexes.
By the way, these results are consistent through time and are reproducible, so it’s not a metter of statistic collector (I guess).
Your observation is the same done at
dba.stackexchange.com and this make me think that the built-in Postgres of OS X Server is
truly optimized.
Best regards,
Pietro
PS on the other response I reported both postgresql.conf
Just looking at the 2 B_2 queries, I'm curious as to why is the execution plan different between the 2 machines. Is the optimiser stats updated on both databases?
Regards,
Wei Shan