Comparing query rates (was Re: 9.2.5 db causes high cpu?)

Jim Reid jim at rfc1035.com
Thu Feb 24 07:38:47 UTC 2005


>>>>> "Brad" == Brad Knowles <brad at stop.mail-abuse.org> writes:

    >> I'm just asking that you try to avoid misleading people by
    >> quoting numbers out of context.

    Brad> 	They should be able to get that by reading the intro
    Brad> to the referenced talk.

You didn't reference it! In your original posting there was no pointer
to an on-line copy of your presentation and you just quoted raw
numbers without explaining the context of how they were obtained.
That's why I asked you to be more careful when posting those results.
Just posting the numbers is misleading.

    Brad> 	All of that is detailed in the slides for the talk, as
    Brad> well you know.

Indeed I do. Did everyone else on this list seen those slides or
know where to find them?

    Brad> 	Everyone has a different environment.  If your
    Brad> requirement for quoting numbers is that the test environment
    Brad> has to perfectly match the one that is used for the
    Brad> production for a given reader, then we'd have to run the
    Brad> test on every single machine and possible combination of
    Brad> machines on the planet.

Brad, you're being silly. What I was and am saying is you should not
quote raw performance numbers without putting them in some sort of
context. That allows people to extrapolate from those results by
comparing that test environment to the one they have. Or they could
decide that the test environment/methodology was flawed and discount
the results.


More information about the bind-workers mailing list