you should in any case try to encapsulate your inserting loop with a transaction. you should still be able to do select statements inside, in particular if all is done within one DB connection. transactions do not speed up considerably, they do it dramatically... ;)
also you may give a code example how you do this, since there are enough sql gurus around here that may help to improve further. hth Marcus > > > > ken-33 wrote: >> >> >> >> look at the sql syntax for insert or replace for sqlite. >> >> Also you goal to handle 1 million per minute is probably going to be >> dependant upon your hardware. >> >> For instance throughput greatly increases with disk striping. >> >> Also the faster the RPM of the drive the more transactions can be >> processed. Code it up and find out! >> >> > > I have actually coded it up. The way I am currently doing it is sending > the > data to a function (the data is 3 doubles) and in that function doing a > SELECT to get the data currently in the DB, then updating the data, then > UPDATE or INSERT. The SQL calls are compiled statements with binds, etc. > It is woefully slow. I was kind of hoping that maybe I was missing a step > or just unfamiliar with the best techniques. I know that when adding a lot > of data with the BEGIN and END TRANSACTION things speed up considerably, > but > I can't use it in my case (I don't believe) because the UPDATE depends > upon > the data in the SELECT. > > John > > -- > View this message in context: > http://www.nabble.com/SQLite-Transaction-Rate-and-speed...-tp22379931p22380540.html > Sent from the SQLite mailing list archive at Nabble.com. > > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users