DSXchange

Your terminology is confusing; you refer to routines but your words suggest that they are job activities. If they are routines, then the trigger can be based on the return value being 0 (OK) or otherwise (not OK). If they are job activities, then make the trigger from SeqSA to SeqODS an OK trigger. ...

You are not losing any data when this message appears. All that is happening is that DataStage is being forced to use disk because there is not enough virtual memory available. I agree with Arnd that it should not be a Fatal message. Do check your ulimit settings for the user ID that processes DataS...

kab123 wrote:I guess there are not many people using CDC stage...

Some of us sleep at night, or take the occasional weekend off! The downside is that now I've got nothing to add to what's already been posted on this thread!

I suspect it's using a ODBC connection to establish queries that confirm the SQL quote and delimiter characters, verify table and column names, and so on.

Try creating an ODBC data source name (DSN) with the same name as your DB2 instance (as it's known to your DataStage job).

The ODBC Enterprise stage is, of course, a parallel stage type. In server jobs you have the ODBC stage, which does not exhibit the same memory leak.

Neglected to mention that the Merge stage performs any kind of key-based join between two text files.

Transformer stages don't read rows, except from their input links.

Failed to read row is much more likely to come from a passive stage whose methods the Transformer stage is invoking.

Please advise what stages make up your job design and what the data source looks like that it's failing to read.

It's called /tmp because it's accessed through MKS Toolkit. Is your file system full? Do you have write permission? Do you, indeed, have a folder called \tmp ? (I had to create one to get 7.5x2 to work - it's not the same folder as Temp.)

"Heap allocation failed" means that it couldn't get more memory when it demanded it. So it has spilled to disk, and continued to process, albeit not as fast.

You still need qsrtmngr to be running on the DataStage server. This is analogous to a run time interface, but specifically for QualityStage. Make sure the port number appears in the hosts file at both ends, too.

we were given a unix path of that new server

First question: has DataStage 7.5 been installed on that server, or are you planning to upgrade the existing server?

Given that you have no apparent capacity for coding, I would suggest the following approach. 1. A server job to read the parameter names (as keys) and values into a hashed file. The name of the parameter file can be passed as a job parameter. 2. A User Variables activity in a job sequence that uses ...

1. Don't allow nulls to reach the input of the Modify stage.

2. Don't allow "" to reach the input of the Modify stage.

You will need to be specific about the layout of the parameters file, because that would - obviously - affect how you would do it. There is no "magic bullet" that works for all cases - you will have some parsing to do.

Anything in the Table Name field becomes the FROM clause in the SELECT statement in Generated SQL. You can quite happily put things like #Schema#.Employee T1 LEFT OUTER JOIN #Schema#.EmployeeDependent T2 ON T1.EmployeeID = T2.EmployeeID into the Table Name field. If you use aliases like that, make s...