Best way to convert to jobs

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
friend.kak@gmail.com
Participant
Posts: 28
Joined: Sat May 03, 2008 3:57 am
Location: chennai

Best way to convert to jobs

Post by friend.kak@gmail.com »

Hi,

I am having a 300 lines of SQL(Oracle), which is joining 36 tables altogether in the same SQL.

Ex:

Select c1.a, c2.b
from c1,c2
where c1.a in
( select b1.a
from b1,b2
where b1.a in
(
select a1.a
from a1,a2
where a1.a in
(
.......
)
)
)

, like that , It is something like inline views in oracle kinda. can anyone let me know what is the bestway to implement the similar logic using Datastage.

>
I just want to simulate the same in Datastage instead of pasting big SQL in one Oracle query :)
Please throw your views on this.
Thanks.
friend.kak@gmail.com
- Dev
vasa_dxx
Participant
Posts: 39
Joined: Sun Sep 28, 2008 2:59 am
Contact:

Post by vasa_dxx »

We also had the similar scenario in our project. We observed that using user defined SQL is efficient instead of using multiple stages in a job.
we had a SQL having 17 joins & all joins were used in a single OE Stage. it works.
But if you need to apply some logic after the data sourcing/Joining, please elaborate the same.
Two wrongs don't make a right. But three lefts do.
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

Stick with the 'big sql' that works rather than breaking it all up to be more 'DataStage like'. If it turns out to be 'too long' for the stage, create a view in your DB and source from the view.
-craig

"You can never have too many knives" -- Logan Nine Fingers
kamalshil
Participant
Posts: 179
Joined: Mon Jun 23, 2008 1:19 am

Post by kamalshil »

Firing SQL is best way.
Multiple join and then sorting of data and partioting may take more time.
Post Reply