DSXchange

Hi All,

While running ajob i am getting the following error.

DataStage Job 5550 Phantom 4056
Program "JOB.2006043476.DT.1489159782.TRANS6": Line 150, Exception raised in GCI subroutine:
Access violation.
Attempting to Cleanup after ABORT raised in stage J_Fact_PS_F_STU_RECRT..Build_PS_E_ADM_PRCT_PROG
DataStage Phantom Aborting with @ABORT.CODE = 3

But if i run with array size=1 and transaction size =1 this job will run fine.
Can any one tell me whats the reason behind this error and how its resolved when i use array size=1 and transaction size =1

Thanks in advance
George Sebastian :D

You'd have to provide details of the job for anyone to help. I've seen issues like this selecting CLOB data from Oracle when the array size is greater than one, doing anything like that?

Thanks chulett.

I dont have any CLOB datatype in my job but i have VARACHAR 2000,which i have used in PIVOT stage. I have 2 similar jobs aborting with
the same reason.

1st job abort will happen in the Transformer just after the PIVOT stage
in the
2nd job abort will happen in the Transformer just before the PIVOT stage.

There are no other warning in the log files.

Thanks
George Sebastian

Aborts are always reported by the transformer as it is the active stage. And sometimes you need to leave the array size at 1 for things to work. What database, by the way? What db stage?

The Database is Oracle and we are using DRS stage.

Thanks
George Sebastian

As an experiment, does it work more better if you replace the DRS stage with an OCI stage?

Thanks for Your reply chulett,

Since we have customers having different databases I am afriad that we cannot deliver this fix.

Thanks for Your reply chulett,

Since we have customers having different databases we cannot use the OCI Stage

I said experiment not solution. Wanted to see if the problem was in the DRS stage or the data itself regardless of the stage of choice.

Thanks for Your reply chulett,
The experiment infact worked with OCI stage. So Is this a BUG with the DRS stage?

Thanks
George Sebastian

Would certainly seem to be. Report the problem to your support provider and see if this is a known issue.

Thanks chulett,

Let me report this issue with IBM and let us see if there is any solution available for this issue.

Will update you all soon

Thanks
George Sebastian [:)]