Freeman Robert - IL wrote: > Thanks Dwayne! Hope you like it! > > It's funny.... after I finally get a copy of a new book in my hands, I > always see something I missed. For example, it just irks me that in my > example of the use of dbms_stats.gather_table_stats that I don't include > cascade=>TRUE in the example. > Don't feel too bad about this one. IMHO, cascade=>FALSE (the default) is the right choice, especially if you use estimate_percent of less than 100%. Then follow with an explicit gather_index_stats with a full compute. And any histograms also with a full compute. -- Regards Wolfgang Breitling Centrex Consulting Corporation www.centrexcc.com -- //www.freelists.org/webpage/oracle-l