Re: Index vs. table scans in statspack reports

"Richard Foote" <richard.foote_at_bigpond.com> wrote:

>If I *cared* about performance, I know what I would *care* most about in my
>environment whether I write the applications or whether each line of (say)
>Java reads like some Egyptian lost tomb encryption.

I fully agree with you, but you don't quite understand my point. Time might be an endless physical magnitude, but my life time isn't, and my working time even less [I've got other responsibilities too]. I still have the hope to engage on the non purely administrative area of tuning, supporting SQL writers. But by now I can't do that, so I feel satisfied if I get my OWN work done.

But remember that SQL tuning is more than just putting an index - you can't tune beyond the schema will allow you to do. In fact, I think the ideal way to tune a database is to adapt the schema and physical layout to the queries, which is the opposite of what is often done. It is rather difficult to know the queries beforehand...

>Rick, learn to care about "missing" indexes ...

As soon as I can!

Bye
Rick Denoire Received on Wed Dec 03 2003 - 17:11:36 CST