Describe failed because of a missing column, Oracle stage

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
harshada
Premium Member
Premium Member
Posts: 92
Joined: Tue May 29, 2007 8:40 am

Describe failed because of a missing column, Oracle stage

Post by harshada »

Hi,

Foll is the job.

Seq file read --> Transformer-->Oracle table load

The column 'FILE_CODE' is not propagated beyond transformer.
Still i am getting the error
"Oracle_Enterprise_15: Describe failed because of a missing column for FILE_CODE".

Write method=load
Write mode=append.

The column 'FILE_CODE' is not present in the table and is of no importance hence not propagated beyond the transformer.

'FILE_CODE' is the first column to be read of the input fixed width file.

I also went to the extent of trying putting the o/p of job1 in a dataset and
then in job2 the dataset -->oracle table.

in job2 there is no mention of 'FILE_CODE'. Still i am getting this error in job2 as well.
Can anyone please help?

Thanks,
Harshada
OddJob
Participant
Posts: 163
Joined: Tue Feb 28, 2006 5:00 am
Location: Sheffield, UK

Post by OddJob »

Have you got Runtime Column Propogation (RCP) enabled for the job/stages?
jreddy
Premium Member
Premium Member
Posts: 202
Joined: Tue Feb 03, 2004 5:09 pm

Post by jreddy »

I had similar problem, but though it doesnt make sense, i added a modify stage inbetween the transformer and Oracle Load, and just a dummy specification like <in_col_name> = <out_col_name>
and it works for me, no more errors.

for eg: in your case, if the column name coming from sequential file is FILE_CODE and in transformer you are mapping that to FCODE that is the column name in oracle table, we usually get the error that you mentioned, that describe failed for FILE_CODE.
My work around is to add a modify between transformer and oracle stage and give just one dummy specification FCODE=FCODE, the input and output metadata for Modify stage is the same as Oracle table.

Try to see if this works for you.
girija
Participant
Posts: 89
Joined: Fri Mar 24, 2006 1:51 pm
Location: Hartford

Post by girija »

I think it occurs due to field name mismatch. Go to the datastage manager. Go to tools---> dataset management.
open the dataset and click on show schema...
Copy the schema definition and search for FILE_CODE.
If it not there change the output column name in your transformer as FILE_CODE.
harshada
Premium Member
Premium Member
Posts: 92
Joined: Tue May 29, 2007 8:40 am

Post by harshada »

Hi,

Runtime Column Propagation (RCP) was a problem. I removed the check and its working now. Thanks for that

Others thanks for the solution provided.

-
Harshada
dvpawankumar
Participant
Posts: 46
Joined: Fri Oct 06, 2006 6:36 am
Location: Tucson

Post by dvpawankumar »

Hi,

It is wierd...I also encountered same error once and I did not do anything. I have deleted the target satge and pulled again from pallate. It worked fine now. RCP was off for me from the begining.

Thanks and Regards,
Pavan.
Thanks and Regards,
Pavan.
Post Reply