Re: Thoughts on SQL tuning disorder
Date: Thu, 20 Jun 2013 15:22:31 -0500
Message-ID: <CAJvnOJafJBgYWoCC8jjpth_3YUM5VMu755xkLiroEWe3SbErng_at_mail.gmail.com>
Its not an easy fix. I would just keep a list of queries that look like they could be improved, and when the subject comes up, present them. Your company is evincing one of the classic disorders I have seen before(I dont know if it has a name), but the symptom is believing and acting on something a consultant says, even though you have been saying the same thing for the last 5 years, and been ignored. When the consultant comes back in a year or so, he will probably identify problem SQL's again...
I have always believe that tuning is an ongoing process, and it never hurts to review SQL performance on a regular basis to watch for potential problems.
On Thu, Jun 20, 2013 at 3:10 PM, <Christopher.Taylor2_at_parallon.net> wrote:
> Cary Milsap covers this idea pretty well in his book "Optimizing Oracle
> Performance" (read it if you haven't as it's an easy read thru most of it).
>
> The idea in a nutshell is if the business isn't complaining about it, or
> if the execution times fall within requirements, then don't spend time
> trying to improve it.
>
> The reason for this is, I assume you have other items on a list of sorts
> you can work on - if you spend time working on something that runs within
> requirements (however they're given - formal or informal) then you probably
> are missing out on accomplishing something else of [more] worth to the
> company.
>
> <SNIP>
>
-- Andrew W. Kerber 'If at first you dont succeed, dont take up skydiving.' -- http://www.freelists.org/webpage/oracle-lReceived on Thu Jun 20 2013 - 22:22:31 CEST