Page 1 of 1

Error while executing parallel job

Posted: Thu Oct 28, 2004 11:48 am
by yaminids
Hello there,

This is the first time I am executing a Parallel job. I am getting the following error. Can any one help me in correcting the error.

##E TBLD 000000 10:13:57(000) <main_program> Error when checking composite operator: Subprocess command failed with exit status 65280
##E TFSR 000019 10:13:57(001) <main_program> Could not check all operators because of previous error(s)
##W TFCP 000000 10:13:57(002) <transform> Error when checking composite operator: The number of reject datasets "0"is less than the number of input datasets "1".
##W TBLD 000000 10:13:57(003) <main_program> Error when checking composite operator: Output from subprocess: Couldn't change directory to /export/Conversion/eCommerce/Customer/extracts:/export/Conversion/eCommerce/Customer/logs: : No such file or directory


Thanks in advance.
-Yamini
[/b]

Posted: Thu Oct 28, 2004 12:06 pm
by gh_amitava
Hi,

Give some more information about your job design and error log in details including Warning.

Regards
Amitava

Error while executing parallel job

Posted: Thu Oct 28, 2004 12:24 pm
by yaminids
Hi,

The design of the job is simple.
SEQ ==> Transformer ==> SEQ.

I got the following error:

Output from transformer compilation follows:

##I TFCN 000001 10:11:29(000) <main_program>
DataStage XE Parallel Extender V7.0.1
Copyright (C) 2003, 1997 - 2002 Ascential Software, Inc.
All Rights Reserved

##I TUTL 000031 10:11:29(001) <main_program> The open files limit is 100; raising to 1024.
##I TOSH 000002 10:11:30(000) <main_program> orchgeneral: loaded
##I TOSH 000002 10:11:30(001) <main_program> orchsort: loaded
##I TOSH 000002 10:11:30(002) <main_program> orchstats: loaded
##I TFSC 000001 10:11:30(005) <main_program> APT configuration file: /opt/dsadm/Ascential/DataStage/Configurations/default.apt
##W TFPA 000017 10:11:30(006) <main_program> Warning: the value of the PWD environment variable does not appear to be a synonym for the current working directory. The current working directory will be used, but if your ORCHESTRATE job does not start up correctly, you should set your PWD environment variable to a value that will work on all nodes of your system.
##E TBLD 000000 10:11:32(000) <main_program> Error when checking composite operator: Subprocess command failed with exit status 65280
##E TFSR 000019 10:11:32(001) <main_program> Could not check all operators because of previous error(s)
##W TFCP 000000 10:11:32(002) <transform> Error when checking composite operator: The number of reject datasets "0"is less than the number of input datasets "1".
##W TBLD 000000 10:11:32(003) <main_program> Error when checking composite operator: Output from subprocess: Couldn't change directory to /xxx/yyy/zzz/aaa/bbb:/xxx/yyy/zzz/aaa/logs: : No such file or directory

##I TFCP 000008 10:11:32(004) <transform> Error when checking composite operator: /opt/SUNWspro/bin/CC -KPIC -O -I/opt/dsadm/Ascential/DataStage/PXEngine/include -dalign -O -PIC -library=iostream -c

Can you help me in correcting the erros.

Thank you.
-Yamini

Posted: Thu Oct 28, 2004 3:06 pm
by ray.wurlod
Always correct the first error first. In your case, this is the error regarding the PWD environment variable. There is another thread updated as recently as today on this topic.

Posted: Mon Nov 15, 2004 12:35 pm
by priyaram
Hi

If ur using the parallel transformer.
and u have any query statements or any statements inside the transformer,just remove all the line feed characters
then compile ur job

Posted: Mon Nov 15, 2004 11:25 pm
by T42
Behind the scene, when DataStage compile a transformer stage, they actually use your specified C++ compiler.

Code: Select all

/opt/SUNWspro/bin/CC -KPIC -O -I/opt/dsadm/Ascential/DataStage/PXEngine/include -dalign -O -PIC -library=iostream -c
If your paths are wrong, then your compiler simply will not work, and therefore DataStage will complain and whine like a 6 month old baby without a pacifier.

Fix that PWD error. If the error continues, make sure the provided paths are established correctly as per your UNIX/Linux rules.

Please let us know of your progress on this.

Posted: Tue Nov 16, 2004 8:39 am
by Eric
To fix the PWD error, you should start the DataStage Engine under a sh shell rather than a ksh shell. :)

Posted: Tue Dec 07, 2004 5:19 am
by aesguerra
Hi,

Are you by any chance making use of custom routines in your transformer jobs?

Posted: Tue Dec 07, 2004 6:07 am
by mandyli
Hi

You are using sequential file stage or Dataset stage.


Thanks
man

Error while executing parallel job

Posted: Tue Dec 07, 2004 1:01 pm
by yaminids
Hello Man,

We are using Sequential File Stage.
-Yamini