DSXchange

Ray, the trace file is quite large so I was just trying to keep the size of the post down. Please let me know if I have extracted enough of it this time. This is just more of Exert1. I suspect it's the same problem, but lets make sure I'm extracting enough of the trace before I post the 2nd exert. 0...

I am trying to understand a particular section in a server-side trace file. The 2 exerts are as follows. Exert 1: 0 Arg4=pExecJobName=RunBatchJC.ETL_DEPRE_LOCN_TO_CC_SEQ pParameterFile=/etl/QA2/parameters/datastage_master_parameters.txt pParameterList=pBATCH_NAME=ETL_DEPRE_LOCN_TO_CC_SEQ|pJOB_NAME=E...

I suppose I should have taken a step back and not gone into solutions mode :oops: So now I'm running a sequence using DSJOB but I've changed a UserStatus of a job within that sequence. If I use the -userstatus option I pressume I'll get the UserStatus of the sequence and not the job? Is this correct...

After a bit of lateral thinking...

How would I output the UserStatus of a job in a unix script?

Arnd,

I believe that would be 'DSJ.USERSTATUS' but that's not what I'm after.

I want to force the JOB status not the UserStatus so that the third party controlling tool gets a different return code from the unix script that runs dsjob.

Does that make more sense?

I am trying to set a special JobStatus for jobs which reject rows based on certain business rules. The job is invoked using dsjob run from a third party scheduler and support needs to know when this situation occurs. I am using a Transform Routine invoked in a derivation of a Reject link. If we get ...

This idea might be a bit like using sledge hammer for a nail but it sounds like something has corrupted on the server.

Re-installing Oracle, then if no luck Re-installing Windows, then Oracle.

Do you have the capacity/security to do this?

Well ideally you could use 1 transformer and perform both lookups in the same transformer using 2 hashed files as reference inputs. Note: The derivation for the 2nd lookup will show up as an error (in red) but the job will compile successfully and run. Try it out. Does anyone know if this is somethi...

Hi Adamski,

Sounds like a security issue to me. Can you manually create a file in that directory using the same user account as the datastage job?

You can obviously do this a few ways. I would personally use Semaphores for locking and unlocking processes you need shared.

This post will help you get started: <a href="viewtopic.php?t=84481&highlight=semaphores">Click here.</a>

I think Datastage uses UTF8 internally as it's own character map. Other NSL character maps store characters with 2 bytes as opposed to 1. I had a similar problem with the TIS620 Thai Character map and record sizes being to small. Whichever field contained Thai characters was doubled and then it stop...

If your license limit is 2, then at any point in time only 2 Datastage clients (Manager, Director, Designer) may be open. If it keeps giving you inconsitent messages just restart the datastage services or bounce the server.

You could try creating the hash files in seperate jobs and potentially firing them off in parallel could be faster. I think leaving the lookups and the tranforms together in the job should be ok (depending on your row size). Landing to disk is always a last resort. It seems your data is small enough...

A few questions for clarification:

Are you creating the Hash files in the same job?

Are your two large bunches of lookups and the transform in the same job?

Are these all seperate jobs and you are landing to disk in between the jobs?

How big is your data?

I think if you start trying NOT to use Nested Condtion or Routine Activity stages in a sequence job you might be limiting yourself a little...

If you client is crashing it sounds to me like your client installtion may be corrupt or you are on a pc with low memory/disk space.