List:General Discussion« Previous MessageNext Message »
From:Michael Widenius Date:November 16 1999 1:31pm
Subject:dual processor machine shockingly slow
View as plain text  
Hi!

Sinisa has already answerd this, but I just wanted to say that to get
better test times for MySQL, you should definitely set up the
key_buffer to a large value (The default is for quite small systems)

Did you find out why your Dual Pentium machine was slow?
I have a similar machine here, so if you want you can always mail me
the script so that I can test it here...

Regards,
Monty

>>>>> "Doug" == Doug Robinson <dkr@stripped> writes:

Doug> Hi
Doug> 	My company is interested in a new DBMS for a
Doug> particular job so I ran a few VERY simple tests on
Doug> various machines. The raw data, such that it is, 
Doug> resides in the attached file.
Doug> 	The tests are about a single table that
Doug> looks very much like the passwd file of a Unix
Doug> system. There is a single index on the login name.
Doug> 	The most shocking result shows up comparing the
Doug> last machine with the first. The first machine is
Doug> a 300MH PII the last a 2x450MH PII. Both run RH 6.0
Doug> Linux. The elapsed times for the dual processor
Doug> are about 10 times those of the single processor.
Doug> 	Does anybody have any explaination for this?!
Doug> dkr

<cut>
Thread
dual processor machine shockingly slowDoug Robinson28 Oct
  • Re: dual processor machine shockingly slowsinisa29 Oct
  • Re: dual processor machine shockingly slowDoug Robinson29 Oct
    • Re: dual processor machine shockingly slowsinisa29 Oct
    • Re: dual processor machine shockingly slowWesley Darlington29 Oct
  • dual processor machine shockingly slowMichael Widenius18 Nov