DSJOB, AWK, AND SUN SOLARIS ...

[sangi1981]

Hi all,
I know that this topic may be off topic, however I hope to find an answer or a suggestion.

I'm writing some shell scripts that launch job with dsjob and do some other things like collect job log (with dsjob -jobdetails ....), and load them on DB.
So, these scripts call other sub scripts that do atomic operation, for example they extract logs and format results.
Particularly these script, after finished execution of main dsjob, invoke another one that calls another dsjob.
When I launch this script from shell, it finish correctly. When I launch this script from a sequence DS, I obtain an error like:

Code: Select all

awk record  '/opt/IBM/Information...'  too much long

[/b]

I don't understand why awk try to process /opt/IBM/..., indeed I use awk only to process string from file.

This is structure of my script:

Code: Select all

main_script
 |
 |_> dsjob -run ... etc ... to run main DS sequences, cores of my process.
 |
 |_> check for status: if status is not ERROR I go on.
 |
 |_> exec subscript that collect log and store on db
       |
       |_> retrieve some info from file that contains name of link, stage: one per job, using, cut, awk, etc
       |_> dsjob -jobdetails .... etc .... etc, for each core sequence
       |_> after collecting all info, run a new sequence that store data on db.
       |_>STOP
|_> STOP
       

It may seem complicated, but the reason was to try to make the code more reusable as possible.

I do not understand the reason for that error. There are too nested calls?
It may be something related to the Sun server?
I know it's difficult to answer a question without being clear about the context, I should post all the code, but I ask to see if the problem may be due to the Sun server.

Thanks in advance for any suggestion