DSXchange

I need to assign a date to a job parameter so it can be used to restrict the records returned from my source system. The date I need will be extracted from SAP BW prior to the job being run so the job knows the last successful load date. It can be extracted into either a hashed file or a flat file....

larryoceanview wrote:I am suprised nobody chewed you out for not searching the forum first.

There's a "honeymoon period" for new members. :D
We do look at the number of posts under the author's identity.
And, no, there's no mechanism for resetting it!

Can you do something like the following:

Which column is the key column?

It is illegal to try to insert NULL into a primary key column. This is true in all databases. Hashed files are the method that the UniVerse database implements its tables. For UniVerse you can also read DataStage Engine.

Yes, resize (ultimately) is a program executed at the operating system level and, like all good UNIX commands (yes, I saw that you're on Windows, but the story's the same), it is silent if successful. There are a few ways to tell whether the file is 64-bit enabled. Probably the easiest is to try UVF...

I think the solution what you have given is difficult to implement I suppose because unless you provide data definition your job doesnt get compiled. Please correct me if I am wrong. - Girish Oak My job design uses only one column. Its name and data type are largely irrelevant, though VarChar is sa...

You could have used ExecTCL as your before/after subroutine and placed the GRANT statement in the Input Value field.
What you did is not wrong; ExecTCL is an alternative.

Typing in partial dsjob commands yields syntax (usage) summaries.

For example (once the DS Engine bin directory is in your PATH), try:

I have a time column which looks like 0814 i need it to look like 08:14 i have tried to use the following code: oconv(iconv(DSLink3.TransTime, "THM[2,2]"), "T:HM[2,2]") I dont get an error nor do I get the any data. It leaves the column null. I am importing it into a sequencial stage where the data...

You'll need to know the internal 'number' of your job, which you can find with a query against DS_JOBS. Easier option: Go to the Project folder on the server, and do an ls -altr and pull the latest set of folders (or the folders with the same timestamp as when you compiled the job.) That's your num...

I think it's worth clarifying the reasons for the unanimity. First, there's nothing technically wrong with using the ODBC stage to perform reference lookups. It will work. But it will work very slowly, leading probably to unacceptable throughput. Pre-loading hashed files and performing the reference...

:D Welcome to the Forum! What you're seeking to do can't be done with a hashed file stage, which uses a computation (called a hashing algorithm) to determine the location of a record based solely on its primary key value. A search of the Forum using the keywords "hashed file" and BETWEEN yielded qui...

Thanks Ray for your reply. Few things probably missed out are as follows : 1.Table definations are stored in my repository. 2. Source Tables and Target have same metadata. 3. I want to create a single job and want to run that with different instances. I want to attach this meta data to the stage on...

Correlatives only ever do integer math. If you want decimal places, apply a conversion, for example (MD2), as part of the expression. It's all in the BASIC manual chapter on conversions and correlatives, not really much to add. If I had my laptop with me I could dredge out a couple of examples. Here...

XMLWriterRUN is the executable for the XML Writer stage. The bad news is that you've found a bug. The good news is that it isn't your fault! Log a call reporting the bug with your support provider. They may require an export of your job design and some sample data in order to try to reproduce the pr...