List:Internals« Previous MessageNext Message »
From:Paul Cadach Date:August 21 2001 1:13pm
Subject:Re: Missed optimization in 3.23.33
View as plain text  
Hi,

----- Original Message -----
From: "Michael Widenius" <monty@stripped>
To: "Paul Cadach" <paul@stripped>
Cc: <internals@stripped>
Sent: Tuesday, August 21, 2001 7:55 PM
Subject: Missed optimization in 3.23.33


> Paul> May be this optimization requres to bring weighting for functions,
to
> Paul> execute simplest functions first, then (if condition is true) more
complex,
> Paul> and last (if condition still true) - most complex functions? This
weight may
> Paul> be a value of time which each function executed 10000 (for example)
times.
>
> The problem is that it's almost impossible for the optimizer to say
> which function will be faster, as this will depend on the arguments
> and the data.  When executing expression MySQL does most things from
> left to right, so it's relatively easy for a human to get MySQL to do
> the right thing.

The problem was solved by using of weightings for comparsion functions at
application level when SQL expression builds.


WBR,
Paul.

Thread
Missed optimization in 3.23.33Paul Cadach20 Aug
  • Re: Missed optimization in 3.23.33Sinisa Milivojevic21 Aug
  • Missed optimization in 3.23.33Michael Widenius21 Aug
  • Re: Missed optimization in 3.23.33Paul Cadach21 Aug