DSXchange

run a job in ds director ,then it shows the info:Job Control Process (pid=xxx) failed

what's wrong ?

That's an easy one - It has failed.

Without supplying more information there is no additional analysis possible.

This is like me asking you what is wrong with my motorcycle because I have a blinking light in the console.

ArndW wrote:That's an easy one - It has failed.

Without supplying more information there is no additional analysis possible.

This is like me asking you what is wrong with my motorcycle because I have a bl ...

there are only 3 lines log
1 control starting job xxx
2 warning job control process (pid =xxx) failed
3 control job xxx aborted

Can you provide few details of your job. Just by seeing your message we cannot analyse anything.

the job is definitly correct, i run it on other servers for many time

today , i import the job into another server. i think there is something wrong in installing or configuration of the new server , but it is hard to tell the every details of the enviroments .

i have googled and got nothing useful .

i think maybe someone have encount this problem.

it seems the job had not been called , and the job didn't run .
the only useful info maybe :Job Control Process (pid=xxx) failed

You get this error when the Job Process was killed.

It might be that some one issued a kill -9 command on your phantom process
or it could be that the process was terminated abruptly due to server errors
or it could be that you have done your node configuration correctly

This got to do with your new enviorment (server).

Did you compile your job after importing it?

If a job control process has failed, you might try resetting it in Director, in case any additional diagnostic information "from previous run" is logged as a result. It may, of course, be an operating system limit that has been reached; your system administrator ought to be monitoring things like the process table and so may be able to advise.

i reinstall the servers, now the job can work.

reinstallation is not a good idea ,but it takes effect.

i still have no idea of what happens to the job ,but it is ok now .

thanks everyone!

my teamate said he had encounted this problem before .
his solution seem quite simple:
1. uv -admin -stop
2. active dsenv
3. uv -admin -start
but it doesnot works in my enviroments.

finally,we found the answer today.

we set CLASSPATH=XXX for java applicition in .bash_profile,maybe the datastage used the wrong classes due to our setting.

if i don't input 'unset CLASSPATH' before 'uv -admin -start' ,the datastage will not work correctly.

the error of " Job Control Process (pid=xxx) failed" will happen !