Page 1 of 1

Sort Stage

Posted: Sat Mar 27, 2010 1:48 pm
by mallikharjuna
Hi,

I have to join(left outer join) the data based on two columns. i have used join stage.my source and reference datasets having huge volumes of data. i have used sort stage before doing the join in both source and reference data based on join keys.

my question is which is the best option, with out sort or with sort stage?


Thanks in Advance
Mallikharjuna

Posted: Sun Mar 28, 2010 12:39 am
by ray.wurlod
The Join stage mandates sorted input, therefore the option of "without sort" is moot.

I prefer to use the Sort stage to using input link sorting, because there is greater control and flexibility over resource usage.