Re: Subquery partition pruning

 > I believe your interpretation of the line 6 filter is incorrect.  Oracle doesn't
> need the "date_wid" comparison at that point because it has already used
> it in the "partition range multicolumn (key/key) in line 4".

Jonathan,

but how should the "PARTITION RANGE MULTI-COLUMN" operation in line 4 pick up the A.DATE_EID without a bloom filter (from 11g on) or a recursive subquery in case of a hash join? My understanding of the "PARTITION RANGE MULTI-COLUMN" operation (introduced in 10.2) is in fact to visualize that only a non-leading part of the partition key is used to prune.

Regards,
Randolf

Oracle related stuff blog:
http://oracle-randolf.blogspot.com/

Co-author of the "OakTable Expert Oracle Practices" book:
http://www.apress.com/book/view/1430226684
http://www.amazon.com/Expert-Oracle-Practices-Database-Administration/dp/1430226684