DSXchange

Sainath's contribution was a UNIX shell script, which means that you would not use DataStage at all for this particular task. The cut command produces a stream containing the first field from each line in the file, where the field delimiter character is "-". This is piped into sort -u which eliminat...

Insert some other Activities to give you the control. Use Routine, Execute Command or Nested Condition Activities - you can effect your control with any of them. For example, an Execute Command activity could test for the existence of a file (you could even use a Wait For File activity here).

There are also all the C-callable functions in the ICI (ic$open, ic$read, ic$write, and so on). Haven't checked whether ICI still ships with DS, it's just an idea. Otherwise you definitely can use the InterCall functions or the UCI functions.

Slams lid back on can of worms.

If I were doing this design I would perform the transformations in DataStage using a Transformer stage, before invoking QualityStage via one, or a series of, QualityStage stage(s). As far as replacing ">" with "&GT" and "<" with "&LT", it appears that you're trying to create HTML tags, but y...

For those on earlier versions, there's a tool called CompileAll available from the good folks who run this site. For FREE! And, before that, was CompileWitch from Wolfgang Huerter - search the archives for more information. Don't know if that's still available.

Unfortunately you won't find the Comments tab in version 6.

Most query languages call this a COALESCE function. It's part of the query language, rather than part of ETL. If your database does not have COALESCE, it would be easy to create as a DataStage routine; I vaguely recall that one was posted either here or on ADN not too long ago. The main difficulty i...

To get a definitive answer we also need to incarcerate Schrodinger's cat in the barrel, and request the answer from the cat. Perhaps Heisenberg could also be engaged as a consultant. Ultimately everything's about quantum, even chaos.

Row-based rules are very difficult to implement, particularly in a parallel environment. If you're not concerned about which row goes into which file, then it's probably easier to do.

It may be easier to use UNIX utilities after PX has produced one file per node to achieve what you require.

Tru64 is an exception to many, many rules!

Craig's recollection is correct, but even before version 6 there had to be a lot of tweaking inside the code to ensure that "word" boundaries were based upon eight-byte words, particularly in shared memory segments.

You can use job parameters for constants in the WHERE clause of the SELECT statement.
Fill in the WHERE clause, without the word WHERE, on the Selection tab of the ODBC stage Output links properties.

Currently DataStage is a 32-bit application. You need to access data through 32-bit drivers for the relevant database.

It's easy (version 7.5) to propagate the Trim() function across all columns in the Transformer stage. Right mouse click and choose propagate expression, then fill in Trim($1) and select "entire expression".

Comparison against @NULL is meaningless; the result is necessarily "unknown". That is, you can't legitimately say Col = @NULL and expect it to behave in a Boolean fashion (in fact, there are rules in DataStage that govern the behaviour under these circumstances, but won't yield the results you expec...

The good news is that your VERSION project is probably OK. Problems with ActiveX occur only on the client. DataStage clients use objects exposed by an ActiveX control (dsobjects.dll or dsobjects.ocx) to communicate with the server. When an object cannot be created it may be that there is insufficien...