Change the design

Post questions here relative to DataStage Server Edition for such areas as Server job design, DS Basic, Routines, Job Sequences, etc.

Moderators: chulett, rschirm, roy

Post Reply
suresh_dsx
Participant
Posts: 160
Joined: Tue May 02, 2006 7:49 am

Change the design

Post by suresh_dsx »

HI all,
we have 100 jobs in my project.
source/ Target --> oracle.
only 30 jobs are we are using source view (DS Quary) in the source oracle stage.
We created one view in the oracle. we splitted in to two views for simplicity.
we are using the second view in the DS Quary.
with in the quary we are joining with the Materlised view from the target table.
How can use that target table directily... If we need to access to the table we need to get DB Links beacuse different schema.
The final aim is to remove the DS Quary in the data stage.. we need to use only one view and also i want to care about tuning the view and also job with less resources.
can you please give some suggestions on this.
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

The best solution will be one that uses no views at all. Second best is to use materialized views.

If you use a standard view you have first to execute the SELECT statement that defines the view before you can even begin to select rows from the view (that is, from the result set of its defining query).

A materialized view only time-shifts this overhead.

If you can select from the base tables, your query constraints can be assisted by indexing. This will always yield the fastest delivery of rows.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Re: Change the design

Post by chulett »

suresh_dsx wrote:How can use that target table directily... If we need to access to the table we need to get DB Links beacuse different schema.
You don't need a dblink to access a different schema - you'd need one to connect to a different instance however. Or to join tables from different instances as if they were in the same instance. Not all that speedily, of course.
-craig

"You can never have too many knives" -- Logan Nine Fingers
Post Reply