DSXchange

Declare a "versioned folder". Place your scripts in there, or place symbolic links to your scripts in there. Version Control can then version the contents of the "versioned folder".

You're in it.

There's another one at Ascential Developer Net

Why can't you use a BASIC Transformer? It's there, it's part of your tool kit.

Stage variable is a good approach - update it only if @INROWNUM = 1.

1. No.
2. No. Unless you uninstall it or rename it.
3. No.

Even at 10,000 rows per second, 80 million rows will take (80,000,000 / 10,000), or 8,000 seconds. You may need to manage expectations.

Welcome aboard. :D The best method, by far, is to write code that doesn't need debugging. DataStage is primarily a graphical tool, and not really set up for debugging in that you don't write the "main" programs. Therefore, in order to be able to use RAID, you would also need to create testbed progra...

More information required. You probably don't need to do anything, because there are no data types within server jobs. If your ASCII value is "3437.34" it can directly be used as a decimal value.

Can't you just use head +1 in the filter pipeline for file2 and file3 ?

Designer can only be connected to one project at a time. However you can have multiple Designers open on one client machine, and these can be connected to different projects. Director can only connect to one project, but can switch to another project. You can have multiple Directors open. Manager ca...

Prefix each row in each file with an artificial key, then use a Merge stage to perform a Full Outer Join.

You can select floor() from the expression editor in the parallel Transformer stage. It's the parallel equivalent of the server Int() function, returning the largest integer smaller than its argument.

... or you could have piped the output of cat through sed to get rid of the linefeed characters

DataStage but Oracle doesn't. When DataStage asks Oracle what the SQL data type is, Oracle reports "timestamp". A DATE column in Oracle is actually a date and time column, as you may be aware.

To convert the date part you therefore first need to extract the date part from the timestamp.

In a Transformer stage you have a delightful tool that allows you to modify the expression in all selected fields. I can't remember the name of the technique right now. If you select one or more output columns and right click any one of them you will find it on the menu. For example, Trim($1) will b...