Fatal Error While loading THrough MultiLOad Plugin

[nivas]

HI

I am trying to load data through TeradataMultiload Plugin.
Inside the job I am calling Server side routine which get the max id from the column. I am incrementing the Id value in the Datastage job.
I am getting the following error.

trn,0: dspipe_wait(15030): Writer(11871) process has terminated.

trn,0: Operator's runLocally() failed

Could anybody please explain.

thanks
Nivas

[T42]

Something failed to work.

Please provide more details. Especially the warning messages before this fatal errors.

Fatal Errors in EE are not the most informative kind of errors. It is the type of errors similiar to listening to a dying man making his final wish. His wish would never make sense unless you actually know what he was talking about.

Don't focus on the last 10 seconds. Be like the detective, and investigate the crime scene (read the warning messages).

[richdhan]

Hi Nivas,

In Teradata Multiload stage there are options to mention the report file name and the directory path where the report file should be generated. Mention the report file as table_name.rpt and mention the directory path.

As far as Teradata multiload is concerned Datastage log will be of no help. You would get better inputs from the report file.

HTH
--Rich

Be quick to listen, slow to speak and slow to become angry

[nivas]

Hi

The job is pulling data from Oracle using Oraoci Plugin and loading into Teradata using Multiload plugin.

The job sometimes only runs successful. While the job fails in the table_name.rpt file sometimes it is writting upto before multiload acquisition stage and sometimes it is not writting any thing.
In the datastage job log file the following are happening.

We have 4 nodes.
The following is the Job Log report when the job successful

First Set
1) bldClaimFsInPXSample131(trn).#3.trn: DSD.StageRun Active stage starting, tracemode = 0.
2) bldClaimFsInPXSample131(trn).#1.trn: DSD.StageRun Active stage starting, tracemode = 0.
3) bldClaimFsInPXSample131(trn).#0.trn: DSD.StageRun Active stage starting, tracemode = 0.
4) bldClaimFsInPXSample131(trn).#2.trn: DSD.StageRun Active stage starting, tracemode = 0.

Second Set
1) bldClaimFsInPXSample131(trn).#0.trn: DSD.StageRun Active stage finishing.
2) bldClaimFsInPXSample131(trn).#3.trn: DSD.StageRun Active stage finishing.
3) bldClaimFsInPXSample131(trn).#2.trn: DSD.StageRun Active stage finishing.
4) bldClaimFsInPXSample131(trn).#1.trn: DSD.StageRun Active stage finishing.
Now the multiload starting.
CLAIM_FS_IN,0: Info: CLAIM_FS_IN: MultiLoad process 5032 has started
After this job is successful

THe following is the job log report when the job fails

First Set
1) bldClaimFsInPXSample131(trn).#3.trn: DSD.StageRun Active stage starting, tracemode = 0.
2) bldClaimFsInPXSample131(trn).#1.trn: DSD.StageRun Active stage starting, tracemode = 0.
3) bldClaimFsInPXSample131(trn).#0.trn: DSD.StageRun Active stage starting, tracemode = 0.
4) bldClaimFsInPXSample131(trn).#2.trn: DSD.StageRun Active stage starting, tracemode = 0.
NOw Multiload process started

1) CLAIM_FS_IN,0: Info: CLAIM_FS_IN: MultiLoad process 6628 has started
2) trn,2: Number of columns returned from shared container does not match that expected by the job.
3) trn,2: Operator's runLocally() failed
4) trn,2: Operator terminated abnormally: runLocally did not return APT_StatusOk
5)
What I am saying is when job fails before running the second set multiload starting ahead. I tried with the change of buffering mode in the OraocI Plugin advanced tab.

Could you please help me in this.

Something failed to work.

Please provide more details. Especially the warning messages before this fatal errors.

Fatal Errors in EE are not the most informative kind of errors. It is the type of errors similiar to listening to a dying man making his final wish. His wish would never make sense unless you actually know what he was talking about.

Don't focus on the last 10 seconds. Be like the detective, and investigate the crime scene (read the warning messages).

[T42]

2) trn,2: Number of columns returned from shared container does not match that expected by the job.

This is the precise error you must investigate. For some reason, you are being passed data from the input to the trn stage on node #3 that have less column values than expected.

One thing I could think of is being passed a NULL in a non-nullable field. But that usually would spawn different error messages.

Contact Ascential Support with the exact information you provided us, along with a copy of your job, the copy of the APT_CONFIG_FILE you're using and hopefully they can pinpoint the cause.

Does this problem ever occurs with 2 nodes, 6 nodes, et cetera?

Did you run out of scratch space?

Did the error occurs right away, or only after a long processing time?