It's a simple table, primary key is an integer, with another column which is integer. Although I do begin a transaction, do 1000 updates and then commit.
On Mon, Mar 18, 2013 at 11:36 AM, Dan Kennedy <danielk1...@gmail.com> wrote: > On 03/19/2013 12:28 AM, Rob Turpin wrote: > >> I was running some performance tests for both sqlite3 and sqlite4 and have >> a question about a difference in my numbers. >> >> I'm running these on a Linux machine, so I enabled fdatasync for the >> sqlite3 build. >> >> I'm measuring the number of updates I can perform (updates per second), >> here are the numbers. >> >> sqlite3: >> Updates (CPU): 156250 >> Updates (Clock): 27733.6 >> >> sqlite4: >> Updates (CPU): 46729 >> Updates (Clock): 33132.8 >> >> >> With sqlite3 there's a large difference between the CPU time and wall >> clock >> time. No big deal, that's the I/O to disk. But then I'm wondering why >> the >> difference with sqlite4 is so small? >> > > How large are the transactions? > > ______________________________**_________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-**bin/mailman/listinfo/sqlite-**users<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