DSXchange

I have a strange situation with my datastage parallel jobs and i dont
have any idea how to go about the problem. Following are the details:

There are couple of jobs which are running on our production
server(protected) successfully but the same jobs when imported to a
different server and are run, we get the following error
"main_program: Fatal Error: Source "DC_TRLR_D" is already kept.".

When we imported the job from prodcution server(job is still protected and readonly) and ran, it went through fine. But when we changed the readonly status of the job to not readonly, recompiled and ran the job on the other server it throws the above mentioned error.

DC_TRLR_D is a field used in most of the stages of the job. It is a key
field and is being used in a couple of lookup stages.

Could anyone tell me as in what kind of possible situations we may get the above mentioned error ?

Just for debugging purposes, if you make the job "not readonly" and "not protected" in the new environment and re-run it, do you still see this error?

DSguru2B wrote:Just for debugging purposes, if you make the job "not readonly" and "not protected" in the new environment and re-run it, do you still see this error?

Yes, I still see the error

Yes, right. It works if its readonly and protected but not the other way around. To be really honest, thats odd. Are all the jobs containing that particular column giving you this kind of behaviour or just some jobs? I think you should get in touch with support.

bvishwanathr


Though it is NOT a very good solution but it worked for us.

Do give a try on this

Just rename the job compile and run it that should solve your problem, do let us know if it really works in you case as well.

Sreedhar wrote:bvishwanathr
Just rename the job compile and run it that should solve your problem, do let us know if it really works in you case as well.

No, It doesnot work.

I understood that there is something to do with my key columns in the lookup stages.

We are actually renaming the column name "DC_TRLR_D" to another name and another column "ACCT_D" to "DC_TRLR_D" in the lookup stage.

When we removed this renaming thing and had the same column names it is working fine. But the strange thing is that the same job(with renaming derivations) is running fine in our production environment.

We are trying on removing the renaming thing out of the lookup stage and having it outside.

Have to see if it works fine.

Kindly suggest some ideas related to the above mentioned issues/descriptions.

bvishwanathr wrote: I understood that there is something to do with my key columns in the lookup stages.

We are actually renaming the column name "DC_TRLR_D" to another name and another column "ACCT_D" to "DC_TRLR_D" in the lookup stage.

I have found the workaround for my problem. In the lookup stage I had the same key names as i have in input(i changed the output field names to match the input field names) and added a modify stage after the lookup stage. In the modify stage i have the column renaming specifications, the job is working fine now.

But still i wonder :
1) What is the property which was creating the problem?
1) What is it that which is making the same job to run fine in our production environment?

Any insights would be appreciated.