Inconsistent run times
Posted: Thu Feb 03, 2005 4:54 pm
Hi,
I have a datastage job that uses an OLEDB connection to a database, runs a script in the BEFORE SQL and then splits the data into one of three text files located on the network. The transform for this job is simple as it only checks the value of a column to separate the data into the files. I am testing so have only 2 to 4 rows that are being processed.
However, when I go to run this job, it takes way too long - on average about 45 minutes to an hour. The weird thing is, occassionally, the job will run in about 3 seconds - the way I expect it to.
I've tested the SQL Script used for running the job and it is fine. I've recreated the job. The files are okay. I just don't know what is going on. It really stumps me when it has run fine on occassion. That and my other DataStage jobs run just fine, even the more complicated ones.
Has anyone run into a problem like this and how would you go about to fix it?
Thanks
Michelle
I have a datastage job that uses an OLEDB connection to a database, runs a script in the BEFORE SQL and then splits the data into one of three text files located on the network. The transform for this job is simple as it only checks the value of a column to separate the data into the files. I am testing so have only 2 to 4 rows that are being processed.
However, when I go to run this job, it takes way too long - on average about 45 minutes to an hour. The weird thing is, occassionally, the job will run in about 3 seconds - the way I expect it to.
I've tested the SQL Script used for running the job and it is fine. I've recreated the job. The files are okay. I just don't know what is going on. It really stumps me when it has run fine on occassion. That and my other DataStage jobs run just fine, even the more complicated ones.
Has anyone run into a problem like this and how would you go about to fix it?
Thanks
Michelle