[GE users] exit status = 25 ?

Ruppert dieter_ruppert at siemens.com
Mon Oct 23 13:14:32 BST 2006


Hi, 

>
>In "common/execution_states.h":
>#define SSTATE_AGAIN               25
>
>And if you look at "daemons/shepherd/shepherd.c", you can see that the
>shepherd goes to SSTATE_AGAIN because of prolog/epilog or
>pe_start/pe_stop exiting with "RESCHEDULE_EXIT_STATUS".
>
>#define RESCHEDULE_EXIT_STATUS 99
>
>So, are those jobs with prolog/epilog/pe_start/pe_stop that can exit with 99??
>

Yes, indeed, it is possible that the job scripts (not prolog/epilog scripts etc.)
exit with value 99. I was not aware that this is considered an error condition; 
we use this as "emergency brake" when the script detects that a job should be
rescheduled elsewhere. 

Is this expected to happen only from prolog/epilog scripts?

Anyway, thanks for the explanation.


Regards
Dieter Ruppert



>On 10/23/06, Ruppert <dieter_ruppert at siemens.com> wrote:
>> Hi,
>>
>> occasionally, I see messages like the following in the messages files
>> of our exec hosts:
>>
>> 10/23/2006 11:37:53|execd|ultraxx|E|shepherd of job 698650.1 exited with exit status = 
25
>>
>> Is this status (25) something created by GridEngine or something which is created
>> by the job itself? If this comes from GridEngine, what could that mean?
>>
>> Regards
>> Dieter Ruppert
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
>> For additional commands, e-mail: users-help at gridengine.sunsource.net
>>
>>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
>For additional commands, e-mail: users-help at gridengine.sunsource.net
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list