DSXchange

You say the buildop has been renamed, but what about the entrypoint in the library (or the object) that was created from it?

Check that the job logs are not becoming unwieldy. Generation of error or warning messages for every row can substantially slow processing. Also purge old entries from the &PH& folder in your project.

I have no experience scripting osh directly, but note that the operator generated by a Data Set stage is copy - that is, the persistent Data Set (on disk) is a direct copy of the virtual Data Set from memory. Get into your DataStage PXEngine directory ($APT_ORCHHOME), cd to the include directory, an...

I cant think of any way you can do this dynamically with DataStage. The Data Set stage does not appear to support dynamic control file name.

Again, what is the behaviour of Floor() with a negative argument?

I think Entire is only recommended because it's idiot-proof. The alternative for a Lookup stage is to guarantee that both/all inputs are identically partitioned based upon the "join" key(s). For Join and Merge stages they must also be sorted on these keys. For Merge stages with more than one update ...

In the pathname \\.\pipe\Project-RT_SC2-Test1

\\. is (or should be) your current working directory (your project?)

pipe is a subdirectory therein

Project-RT_SC2-Test1 is the name of a pipe

Theoretically no. Practically yes, because it complains if you don't.

By default any object that ships with DataStage (including the SDK transforms and routines) are marked read only, and are not included in an export on the grounds that they will also be present in any other DataStage project. This - and other strategies such as not exporting any property that has it...

The file name for a local message handler is Local.msh and it lives in the job's RT_SCnnnn directory (where nnnn is the job number from DS_JOBS). Named message handlers are stored centrally.

There IS a document about setting up DB2 connectivity; it doesn't ship with DataStage but has to be requested through your support provider. Maybe if you search the forum you can get its title.

Perhaps you don't purge your logs and this particular one has hit its 2GB limit?

As far as I am aware Key is irrelevant in Data Sets. It is pertinent in Lookup File Sets. Please harangue your IBM rep for additional information.

Select Stage Types branch in Repository. Right click, choose Add New Parallel Stage then, from the next menu, choose Build. Use the F1 or Help buttons within the stage editor to figure out what to do from there; what code to put on which tabs.

Be careful if the number of months to be added is negative.
Be careful also to check that the resulting date is valid (for example 31 Jan + 1 month).