-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 24/10/11 13:23, Volker Jahns wrote: > Any help is greatly appreciated on how to solve this issue.
You have given no way for anyone else to try to repeat the issue. We can't see your source code and it is unlikely your benchmark is the first to have no bugs whatsoever. Two sql scripts, one to populate the databases with identical content and one to run the query would help. It isn't obvious if the timings are from running the query once or from a large number of runs. If the former then you really will get something close to random numbers. In any event this test is completely bogus in many ways. You are altering the processor speed & cores, RAM, operating system, SQLite version, hard drive and other programs that are running for each one. They will all affect the performance of a query. Unless the Dell machine you are running MacOS on (!?) is using an SSD, its timing is nonsense and certainly not repeatable as it is less than the seek time of even current hard drives - ie something is lying about data being on disk platters. Roger -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) iEYEARECAAYFAk6mFEQACgkQmOOfHg372QQZiACgxY5m7iSrOE9JIUC+nvS7UOKu bRsAnicTMrGXWf7RyatOBZA7YZRtW6p+ =yPt8 -----END PGP SIGNATURE----- _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
