Why takes this select AGES (and hogs CPU like ...)

Hello All,

I have created a table (space_values) that contains the out parameters of the dbms_space package. on each day, a new row is created in this table. in a separate table (space_tables) i have the table names and owners of the tables i analyze with dbms_space. Now i want to query the growth of the tablespaces using
my space_values and dba_tables (or all_tables). For that purpose, i wrote this query:
(with hwm=(total_blocks-unused_blocks-1) and datum = date when analyzed
in space_values are columns owner, table_name and datum the primary key

select distinct aa.datum, round(sum(aa.hwm)/256,2) from space_values aa, dba_tables bb
where bb.tablespace_name = 'tablespacename' and bb.table_name = a.table_name
and bb.owner = aa.owner
group by datum

but this takes AGES !! And i don't know why since there are only 800 rows in space_values.
when i remove the group by and the datum in the statement, the query is pretty fast.
The idea was to get the amounts of space for each day and each tablespace with one query, but it
takes all day even when i query
only one tablespace, as you see above. What am I doing wrong?

Any help would be greatly appreciated!

Frank Zaum, Salzgitter Received on Thu Apr 25 2002 - 21:18:22 CEST