On Mon, 06 Oct 2008, Francesco Saverio Giudice wrote: Hi Francesco,
Many thanks for your all tests. I'll try to look at them closer tomorrow. >> Now I'm running same tests on a Windows 2003 Server 32bit with 2 CPU Xeon >> Dual Core 3Ghz (total of 4 CPUs). I will send tests in a while. > look at attached logs. Absolutly *out of mind* results. > Probably due to "server" personality, times are out of ranges. > There are also some tests with * chars because out of ranges. In the 1-st case it was midnight effect and SECONDS() counter resetting. In the second one I do not know what was the reason. Maybe some clock updating? We are using real time clock so such things can happen. This test shows that it consumes 870.36 seconds from all CPUs but in 3927.41 seconds of real time. Looks that sth froze the process execution for nearly an hour (f.e. some other process with high priority or simply system clock was updated). > Tests executed on: > Microsoft Windows Server 2003, Standard Edition Service Pack 2 (Build 3790) > 2 x Intel(R) Xeon(TM) CPU 3.00GHz Dual Core > L2 Cache Size: 2048 Kb > Bus: 200 Mhz Socket > 3,00 Gb Installed Memory (1,19 Gb free) > in a Terminal Session. In this test system load level is very important because we will calculate also time consumed by other processes - side effect of using real time clock. Probably it will be optimal to run this test only in 2 threads excluding test with memory allocation to reduce the interactions with other programs. Once again thanks to you and all others for test results. best regards, Przemek _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
