Performance of datasets

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
go4it
Participant
Posts: 34
Joined: Wed Apr 23, 2008 5:49 am
Location: UK

Performance of datasets

Post by go4it »

Hi,

I have around 15 million records extracted and written to a dataset 1 and transformed the dataset 1 to dataset 2 ( not much changes, mostly one to one mapping), however the issue I have is it takes a while to write to the dataset 2 and takes much longer time to load the data from dataset 2 into the Oracle table.

Could you please advice on what best could be done to improve the performance on this partiuclar job to Extract, Transform and Load the data ?

Would using a Fileset be better opion ?

Pls advice

Mahesh
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

A File Set will always be slower than a Data Set. File Set uses export and import operator, while Data Set uses copy operator.

The slowness loading the Oracle table is something in Oracle. You can prove this by constraining the rows to pass zero rows to Oracle, thereby getting a measure of the raw read speed of a Data Set stage.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
go4it
Participant
Posts: 34
Joined: Wed Apr 23, 2008 5:49 am
Location: UK

Post by go4it »

Thanks ray, but the problem here is when I to write to another datasets.

Dataset1 -> xfm -> Dataset2

Takes more time and aborts.
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

'Takes time'? 'Aborts'? Of course it takes time, you'd need to let us know how much and why you think it's too long. And post your abort message, of course.
Last edited by chulett on Thu Oct 29, 2009 8:02 am, edited 1 time in total.
-craig

"You can never have too many knives" -- Logan Nine Fingers
go4it
Participant
Posts: 34
Joined: Wed Apr 23, 2008 5:49 am
Location: UK

Post by go4it »

Hi Ray/Craig,

I tried teh following and it worked, when running I changed the Config file and it worked fine, but when the next job to load the dataset in oracle was called it was taking a while and aborting after a while.

On discussions with the DBA, found that there was not enough space and the number of redo log files were increased.

Its working fine now.

Thanks for your help.
Post Reply