Oracle Scratchpad

February 17, 2012

RMOUG Debrief

Filed under: Advertisements — Jonathan Lewis @ 8:24 pm BST Feb 17,2012

I’m back home after a very pleasant few days in Denver. As always the RMOUG conference was great fun, the location delightful, and the people were friendly, and the organizers did a great job of looking after the presenters and making things happen. It’s a terrific event – just big enough to have a real buzz, but small enough (I’d guess about 700 – 800 people) that you don’t feel lost in the crowd.

All three of my presentations seemed to go well – one person even started to applaud (which got the whole room going) when I got to the really important slide in the first session.

Jonathan Gennick of Apress had brought 100 copies of Oracle Core and was selling them off at a special event price of USD 20 – and I think he only had one or two copies left by the end of the event; I know that I autographed an easy 40 or 50 copies.

One of the nicest things about the event happened just a few minutes ago, though. Just as I was about to start typing up this little note, I got the following email, which I reproduce with permission:

I just wanted to say I appreciate the time and effort you put into your presentations on table access and joins at RMOUG.  They were excellent, as always.  Based on what you showed us, I have rewritten one problem query in our application such that I am getting an 82 percent performance gain.  I am excited about the possibilities and look forward to sharing what you showed us at RMOUG to our development team.

Thank you again for the support you have shown for RMOUG and Oracle professionals.

Sandy Becker
Database Operations Manager
Oildex, a service of Transzap, Inc.

It’s really nice to hear that someone has taken a mechanism you’ve described and found examples where it can be applied to such good effect and done it so soon after the event.

5 Comments »

  1. Hi Jonathan,
    Thats really awesome. Where can we find the slides :)

    Comment by Raghav — February 17, 2012 @ 10:36 pm BST Feb 17,2012 | Reply

  2. Exactly Raghav. That was aslo my first thought. Then I started to think about it. Jonathan just released his new book “Oracle Core”. He must have shared some of those “secrets” inside this new book :-) I allready got the book, now I just have to keep reading. By the way … this is an excellent book. Very understandable and easy to read. I attended a 5 day class with Jonathan in Bergen, Norway about 4-6 years ago. The compendium from back then started to get really worn out. It was about time that this book was released :-). Jonathan: Are you still giving those 5 day classes?

    Comment by Lasse Jenssen — February 18, 2012 @ 10:43 am BST Feb 18,2012 | Reply

    • Lasse,

      The course was only 3 days – though it may have felt like 5 given the volume and quality of information. The supporting notes from that course totalled about 400 pages, with about 100 SQL scripts. I had to stop presenting it eventually because Oracle was releasing upgrades and patches faster than I could keep the notes up to date.

      Raghav,
      If you’re in the USA, I’m scheduled to do the same three presentations at ODTUG in San Antonio on 24th – 26th June.
      Alternatively, I’ll be doing one of my full day tutorials that covers the material in Minneapolis and Salt Lake City in mid May, and in Dallas at the end of June.

      Comment by Jonathan Lewis — March 12, 2012 @ 2:05 pm BST Mar 12,2012 | Reply

  3. Jonathan,

    These are real comments from our code:

    Special optimizing of the HASH JOIN. The big hash join caused too big TEMP space consumption…..Join ONLY ROWID to avoid joining all the large columns. More on this method here: http://jonathanlewis.wordpress.com/2008/05/09/manual-optimisation-2/

    XML will be generated directly with the Select itself. The old way (appending the rows and columns one after another) prooved to be very slow because of the numerous concatenations of CLOB with strings. See here for further explanation of why it was slow: http://jonathanlewis.wordpress.com/2009/07/09/concatenating-lobs/

    And there are other solutions inspired by you, not explicitely commented:

    – The NO_MERGE hint solved us numerous problems by complex view merging. I can count more than 70 usages of the hint.

    – The NO_UNNEST hint saved our lifes in a specific situation with subqueries.

    – Why is the PARTITION EXCHANGE full scaning the tables – even when *whithout validation* is specified? There is an old but very usefull article by vou on how partition exchange is dealing with constraints – which opened my eyes on the subject.

    I will stop here because the post got rather long – and these are all things which I just thought of the top of my head.

    Comment by Todor Botev — February 20, 2012 @ 10:11 am BST Feb 20,2012 | Reply

  4. Nice to finally talk briefly with you in person!

    Comment by Jason Bucata — February 20, 2012 @ 7:04 pm BST Feb 20,2012 | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Theme: Rubric. Get a free blog at WordPress.com

Follow

Get every new post delivered to your Inbox.

Join 4,258 other followers