Parallel job reports failure (code 139)

[nraj]

Hi,

I have created a simple job which loads data from Teadata to Dataset file,
job got aborted with fatal error,

Parallel job reports failure (code 139)

Plz advice me how to go about it,,,,

Thanks

[ArndW]

nraj,

you really need to offer some more information and also to try to narrow down your source problem yourself. Here are some questions that might get you thinking or testing in the right direction, as the error message is not very helpful:

- does the error occur right away or after the job processes rows?
- if you write to a sequential file instead of a data set does the error persist?
- if you change the number of nodes in your APT_CONFIG does the error change behaviour?
- can you do a 1-1 mapping on another table?
- if you change your select (perhaps via an ORDER BY) does the error change?
- If you reduce the number of columns selected/copied in your database stage does the error go away at some point in time?

[kumar_s]

Gotcha!!!!!!

Got another case like mine!!!! I too breaking with same sort of issue for past 2 days.
But its DB2 in my case and not unload but Load process.
And the code 137 is a SIGKILL.
I use a load script to load a table. I dont find it as an data issue, because the same occurs if i try with empty file. When i try the same script from command propmt it works .
Necesssary privilage was issued to tables and users. Table was droped recreated. All other table goes fine.
I i replace the LOAD command by some other command say SELECT the script pass through. Need to find the reason of SIGKILL while it been called from DataStage

[ray.wurlod]

SIGKILL isn't called from DataStage. This indicates that someone has used kill -9 to kill one or more of your processes.

[kumar_s]

SIGKILL isn't called from DataStage. This indicates that someone has used kill -9 to kill one or more of your processes.

Though none delebrately issue this command, the shell gets kill with this code when called by datastage and not from command prompt