I wouldn't really expect to get any
significant benefit from partition
elimination when using hash partitions.
In fact, apart from the administrative
convenience, I think the only benefit
is likely to come from partition wise
joins using partition iterators.
Can you give us any details of the
You can get significant benefit from hash partitioning if you have 2 or more
tables that are equi-partitioned. The benefit comes when you join the tables
together -- you get partition-wise joining! In effect, you perform multiple
small table joins instead of 1 join of 2 large tables!
This works fo
As we have discovered, hash partitioning does not
always give you partition elimination nor does CBO
always work as expected. One of our DBAs has a tar
open with Oracle on this issue and we are finding that
no one appears to understand the complete
ramifications of using hash partitions when a que
I'm not sure what your trying to accomplish. You get one insert for one
record. Which partition is used depends on how you set them up. If your
wanting the partitions to be used in a circular fashion I believe that is
accomplished with a hash partition and you get to create the hashing method.
YES, here's an example:
create table ate_headers(module_id varchar2(13),
session_number varchar2(16),
test_group number(4),
test_date date,
first_record char(1) default 'F',
John,
First question, how do you know that your application is rule based? Most
application scan be switched to cost optimization with no changes, although I
will admit performance sometimes goes south. I prefer to set the database to
'choose' mode which allows the best of both worlds.
Dic
