On Thu, Feb 9, 2012 at 2:08 AM, Steinar Midtskogen <stei...@latinitas.org>wrote:
> > When I build my database from scratch using millions of inserts, one > table causes problems. Inserts get slower and slower. I have about > 830,000 inserts for that table. It gets to 300,000 pretty fast, but > then it gets slower and slower, and eventually it will only do a few > inserts per second, and I then I have to kill sqlite3 as it will run > for hours if not days. > My guesses: - Your unix_time values are not successive. In this case your first fast results are due to advantages of memory caching. The following slowness is the result of the internal fragmentation - You have successive unix_time, but when it goes beyond 32-bit limit your software wraps around 32-bit value instead of going 64 bit. Max _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
