Comments on: True or False – Improving Performance of SQL Statements http://hoopercharles.wordpress.com/2010/03/30/true-or-false-improving-performance-of-sql-statements/ Miscellaneous Random Oracle Topics: Stop, Think, ... Understand Mon, 13 May 2013 14:10:06 +0000 hourly 1 http://wordpress.com/ By: Charles Hooper http://hoopercharles.wordpress.com/2010/03/30/true-or-false-improving-performance-of-sql-statements/#comment-595 Mon, 05 Apr 2010 19:45:29 +0000 http://hoopercharles.wordpress.com/?p=1807#comment-595 #3: Probably false 99% of the time. Article 5 implies that it is true, article 6 implies that it is false.

#4: Probably false 99% of the time. This suggestion was found in article #9, a point that I mentioned when I wrote a review for that book.

Blue moons and double-headed coins happen on occasion too.

]]> By: Log Buffer #184, a Carnival of the Vanities for DBAs | The Pythian Blog http://hoopercharles.wordpress.com/2010/03/30/true-or-false-improving-performance-of-sql-statements/#comment-577 Thu, 01 Apr 2010 22:32:04 +0000 http://hoopercharles.wordpress.com/?p=1807#comment-577 [...] Hooper posted a 3-part series with seemingly innocent True/False questions. He covers sorting, SQL tuning and wait [...]

]]> By: Charles Hooper http://hoopercharles.wordpress.com/2010/03/30/true-or-false-improving-performance-of-sql-statements/#comment-572 Wed, 31 Mar 2010 10:50:40 +0000 http://hoopercharles.wordpress.com/?p=1807#comment-572 Well stated.

Many of these True/False questions are stated as implied absolutes, so part of the challenge in answering the questions is trying to find conditions that lead to the statement evaluating as false. Take, for instance, question #1. If you follow the documentation link for subqueries, you will find that some of the approaches suggested in the articles for removing subqueries actually generate another form of a subquery. I personally do not consider an inline view to be a subquery, but the Oracle documentation states that it is a subquery. Another possible reason for the question to evaluate to false involves the automatic query rewriting capability in Oracle 9i and above – is it a logical step to always perform the transformation – Oracle 10g and above suggests not.

]]> By: Chris_c http://hoopercharles.wordpress.com/2010/03/30/true-or-false-improving-performance-of-sql-statements/#comment-571 Wed, 31 Mar 2010 10:16:10 +0000 http://hoopercharles.wordpress.com/?p=1807#comment-571 As its more fun than waiting for yet another test cycle to complete:-

Number 2:
As an old friend form Glasgow used to say Maybe is Aye Maybe is Naw.
Usually though I use the with clause or GTT to simplify a complex problem, either because the syntax of the query is horrible or I can’t figure out a single query to get the result I want (this is a pretty good indication of either a bad design or a bad specification). I have used the With clause when performance tuning but usually in combination with something else (analytics in this example http://forums.oracle.com/forums/thread.jspa?messageID=2187453&#2187453) here the design is so bad and the subquery used so often that the cost of writting the factored query to temp then reading it back into the buffer cache is worthwhile.
The other case I have used this for performance is when the optimiser couldn’t find a good plan not because of the subquery but because the data was being pulled from a complex underlying view factoring out the subquery resulted in a better plan, currently I don’t have a test case for it though building one and figuring out exactly what happened is in my todo list so its currently filed under Voodoo tuning methods.

Chris

]]>