|
|
|
Hall of Fame
       
Group: General Forum Members
Last Login: Yesterday @ 2:02 PM
Points: 3,788,
Visits: 5,543
|
|
At the last PASS summit, I had a chance to talk for about an hour at breakfast with a man from Microsoft who is part of the BOL team. According to him, so many new things are coming out so fast that just keeping the documentation up to date is a real challenge.
I thought that the question was a good one, and the discussion even better. (I got it wrong because of my interpretation, but so what?) A lot of information about index structures and index types came out in the discussion.
__________________________________________________
Against stupidity the gods themselves contend in vain. -- Friedrich Schiller Stop, children, what's that sound? -- Stephen Stills
|
|
|
|
|
SSCertifiable
       
Group: General Forum Members
Last Login: Yesterday @ 2:02 PM
Points: 6,367,
Visits: 8,228
|
|
The Dixie Flatline (5/27/2010) At the last PASS summit, I had a chance to talk for about an hour at breakfast with a man from Microsoft who is part of the BOL team. According to him, so many new things are coming out so fast that just keeping the documentation up to date is a real challenge.
I thought that the question was a good one, and the discussion even better. (I got it wrong because of my interpretation, but so what?) A lot of information about index structures and index types came out in the discussion. Thanks.
One thing that I've noticed is that MS doesn't annotate when the pages change. So, in the case of this QotD, several references have changed since I originally submitted the question. I've also noticed that BOL doesn't have links to previous versions like they used to anymore.
Wayne Microsoft Certified Master: SQL Server 2008 If you can't explain to another person how the code that you're copying from the internet works, then DON'T USE IT on a production system! After all, you will be the one supporting it! Links: For better assistance in answering your questions, How to ask a question, Performance Problems, Common date/time routines, CROSS-TABS and PIVOT tables Part 1 & Part 2, Using APPLY Part 1 & Part 2, Splitting Delimited Strings
|
|
|
|
|
SSChampion
        
Group: General Forum Members
Last Login: Wednesday, May 08, 2013 10:33 AM
Points: 10,989,
Visits: 10,529
|
|
WayneS (5/27/2010)
Question for you Paul: It is stated in BOL: Instead of constructing a B-tree structure based on a value stored in a particular row, the Full-Text Engine builds an inverted, stacked, compressed index structure based on individual tokens from the text being indexed. Now, I'm not about to question Kalen, but where MS is stating so specifically that it's not a B-tree, what gives here? Hey Wayne,
This is just wording again - and it is confusing. My guess is that the intention of the writer was to show that full-text indexes have a different internal arrangement from 'ordinary' indexes. The distinction is between a B+ tree based on a value stored in a particular row, and a B+ tree built on an inverted, stacked, compressed structure build on parsed tokens.
If you have Kalen's 2008 Internals book, the relevant sections are on pages 345-346. A key part (talking about FT indexes) is "...their space usage is tracked in exactly the same way (using IAM pages) and their structures are the same as regular indexes."
As I keep saying, though, whether you consider full-text indexes and 'regular' indexes to have the same structure or not depends a lot on what level of structure we are talking about.
Paul
Paul White SQL Server MVP SQLblog.com @SQL_Kiwi
|
|
|
|
|
SSCrazy
      
Group: General Forum Members
Last Login: Monday, May 13, 2013 11:21 AM
Points: 2,163,
Visits: 2,148
|
|
Great questions thanks!
I got it correct, but not for the right reason.
|
|
|
|
|
SSCommitted
      
Group: General Forum Members
Last Login: Tuesday, May 14, 2013 1:23 PM
Points: 1,905,
Visits: 1,601
|
|
|
|
|
|
SSCertifiable
       
Group: General Forum Members
Last Login: Yesterday @ 1:10 PM
Points: 7,084,
Visits: 7,137
|
|
Interesting question and discussion.
I got it wrong (said 4) so I got my point. 
edit: But I'm still not sure whether the right answer is 1 or 2, even after reading all the discussion. I guess it depends on whether a B+tree indexing rows by column values is the same structure as a B+tree indexing some sort of pile of tokens attached to pointers (if that's what a full text index is - I don't really understand that either).
Tom Que conclure à la fin de tous mes longs propos? C'est que les préjugés sont la raison des sots. (Voltaire, 1756)
|
|
|
|
|
Old Hand
      
Group: General Forum Members
Last Login: Monday, February 18, 2013 10:54 PM
Points: 371,
Visits: 110
|
|
got it right... this time with my understanding..........
|
|
|
|