I have encountered one issue wherein an insert job(in the sequence) kept executing for a long time (abt 5 hrs) without actually affecting the table.
Before this insert operation, a delete operation(in the same sequence job) was successfully done on the same table.
I am not sure as to why this insert job (although the DataStage Director showed its status as 'Running') didn't affected the table any way.
Finally, I aborted the job and executed it again and this time it executed successfully within time limits.
Please help me to track what went wrong with the job.
Job not responding in the JobSequence
Moderators: chulett, rschirm, roy
Hello Abby,
from your problem description it is impossible for us to find what has caused your original issue. When a job seems to "hang" and doesn't consume a lot of CPU time it is most often due to locking; either within DataStage or within the database(s). If this happens again you can try to get some more information - do the unix processes use CPU? Does the DBA for the database show any table, group or record level locks on that table? Does the job monitor show any rows processed at all? Without some narrowing down of the possible causes there really is no answer possible.
from your problem description it is impossible for us to find what has caused your original issue. When a job seems to "hang" and doesn't consume a lot of CPU time it is most often due to locking; either within DataStage or within the database(s). If this happens again you can try to get some more information - do the unix processes use CPU? Does the DBA for the database show any table, group or record level locks on that table? Does the job monitor show any rows processed at all? Without some narrowing down of the possible causes there really is no answer possible.
</a>Hello ArndW,
Thanks very much for the inputs!!!
Well, as far as locks are concerned there were no locks on any of the tables in use. Also, the Director logs does not provide any information on the number of rows processed. Apart from this, there are no capture files created against the required user_name on that particular date.
But, this problem is not specific to a single job. Next time, I found another job hanging for long hours. Again, we aborted the job for manual execution. It might be possible that the job hanging is related with some environment issue. I'm still chasing the cause of the problem.
Thanks very much for the inputs!!!
Well, as far as locks are concerned there were no locks on any of the tables in use. Also, the Director logs does not provide any information on the number of rows processed. Apart from this, there are no capture files created against the required user_name on that particular date.
But, this problem is not specific to a single job. Next time, I found another job hanging for long hours. Again, we aborted the job for manual execution. It might be possible that the job hanging is related with some environment issue. I'm still chasing the cause of the problem.