RE: clarification on partitioned tables

I don't see the parameter 'granularity' set. The default is 'ALL' and since cascaded is set to true it means that for event if you set partname you will be caluculating any global index again and again for each partitions, subpartition. Cheer !  

 Beside that, with so much indexes, I doubt your TEMP even set to something to 60 to 100GB will make it. You gonna get unable to extent tmp after 2 or 3 days.  

"

 exec dbms_stats.gather_table_stats (ownname=>'schema_owner',tabname=>'table_name',  partname=>'partition_name',estimate_percent=>DBMS_STATS.AUTO_SAMPLE_SIZE,  method_opt=>'for all indexed columns', cascade=>TRUE);
"
   

 B. Polarski
 http://www.smenu.org

--
http://www.freelists.org/webpage/oracle-l