INT function in 7.5x2 vs 8.0.1

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
TD_Chile
Premium Member
Premium Member
Posts: 10
Joined: Tue Jun 10, 2008 1:43 pm
Location: Santiago, Chile

INT function in 7.5x2 vs 8.0.1

Post by TD_Chile »

We're migrating from 7.5x2 to 8.0.1 and when doing the tests we found this problem: the INT () function doesn't return the same value!!! maybe because the value is null but in 7.5x2 a null value in a sequential file it's the same than empty and INT(emptyvalue)=0!!!... but in 8.0.1 a null value in a sequential file is null so INT(nullvalue)=NULL!!!!... We have more than 200 jobs using this function in diferent ways, someone has an idea for what to do?

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

Post by ray.wurlod »

Have you specified different NULL mapping in the job? What is the default null mapping in each version (I don't have 8.0.1 where I am)?
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
TD_Chile
Premium Member
Premium Member
Posts: 10
Joined: Tue Jun 10, 2008 1:43 pm
Location: Santiago, Chile

Post by TD_Chile »

I don't know if I can specified a different NULL mapping, if yes that could affect the hash files in wich a NULL always it's a NULL, no?
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

In hashed file NULL is always NULL. Internally in hashed files NULL is represented as Char(128), but the translation is automatic. It is only in text files that you need to specify how NULL is represented. My question was whether the default handling is the same in version 8.0.1 as it is in 7.5x2, because I am not presently in a position to verify that for myself.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
TD_Chile
Premium Member
Premium Member
Posts: 10
Joined: Tue Jun 10, 2008 1:43 pm
Location: Santiago, Chile

Post by TD_Chile »

well, after doing a lot of tests the conclusion is that null values on DS7 are treated different (clearly it's a bug) and in DS8 null values are always treated as null, in both sequential files and hashed files.
I also checked the NULL mapping in the sequential files but it's the same, so all that we built assuming "empty = null" on seq files must be changed

thanks
Post Reply