Sidenote: The post above refers to DB2 for z/OS which has a different
implementation than DB2 for Unix et. Al.
DB2 for Unix also has clustered indexes, but they are on a "best effort"
basis.
I'm not a DBA, but I could imagine that row-chaining would be a common
req'd to do reorgs.
How much of that can be prevented by design (other than by avoiding
VARCHAR() ;-) I do not know.
Come to think of it, recovering extends (collections of pages) after a
table shrinks, might be another reason for reorg in DB2 for Unix...
Cheers
Serge
--
Serge Rielau
DB2 UDB SQL Compiler Development
IBM Software Lab, Toronto
Try the DB2 UDB V8.1 beta at
http://www-3.ibm.com/software/data/db2/
Received on Mon Oct 07 2002 - 22:41:06 CDT
