> > On 2014/12/11 17:58, Paul wrote: > > > >> On 2014/12/11 13:51, Paul wrote: > >> I have yet to try and test if dropping stat tables worth the effort. Some > >> databases in fact can grow pretty big, up to few > >> hundred of megabytes//.... > > In that case maybe keep the Stat1 tables and there is also the option of > using stuff like "USING" clauses and "LIKELY" or "UNLIKELY" > planner directives in your queries to force a tried and tested query plan on > the QP in lieu of using the stat tables - but now you > are getting very hands-on with your data and no longer leaving it up to the > internal wisdom of SQLite - something I don't usually > advocate, but as you rightly observed - your case is quite special. >
I am going to disable stat4 and keep only stat1. Yes, I could of use USING and co. But as you correctly pointed out, sqlite does very good job optimizing queries. I trust sqlite more than I trust my intuition. >From my personal experience, intuition does not perform very well in complex >systems with a lot of variables. Best regards, Paul _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
