Index Efficiency

I was looking for an old posting on the Oracle newsgroup recently, when I found a set of posts that talked about measuring the effectiveness of rebuilding some indexes.

I had contributed by pointing to my website where I have a script that demonstrates how to get some detailed information about how well the space in an index is used, and Joel Garry had responded with some results from a production instance.

The full thread starts at this URL

But for a quick summary of the conversation I had with Joel, you could go to each of these posts in turn:

• My comments about doing a tree dump of an index
• Introducing my “leaf block scan” code:
• Joel’s repsonse after using the “leaf block scan” code
• My analysis of Joel’s output
• Joel’s response to my analysis

 For other notes I’ve written about index rebuilds:

• A bug in 10g after index rebuilds
• A little-known threat from index rebuilds

Happy Birthday!

This year marks the 25th anniversary of the UK Oracle User Group - and we have a target of 430 presentations covering the many different faces of Oracle’s presence in the IT and business world.

The call for papers has just gone out, with a closing date of May 9th - so if you want to join the fun, sign up soon.

Update:

(1) The closing date has been extended to 20th May - I know it seems early for a conference at the start of December, but there’s a lot of work in putting together an event for 2,500 people and early submissions help a lot.

(2) Since Oracle has just bought BEA, the UKOUG will be putting together a stream for users of their product as well - even if the list of topics in the ‘abstracts’ screen doesn’t include the right label  yet.