DSXchange

It IS possible to write interruptible sleep routines, but it requires operating system calls and therefore would probably be need to be written in C (and, if server routines need to invoke the function, it would also be necessary to create a GCI subroutine definition - you don't really want to go th...

Not here. There are separate fora for TX (WebSphere Transformation Extender), for Information Analyzer (formerly ProfileStage formerly Mercator), and for lots of other things. Explore. However the form in which you posted is for general questions about data integration (stragegies, tools, techniques...

What values do your four APT compiler/linker environment variables have?

It is not possible.

It is documented as not being possible.

Indeed, it would be easy to code, and you could put differential sleeps into a routine invoked from a Routine activity.

Trivium: "Apt" as a word is the antonym of "inept".

Check the UNIX kernel limit on the number of processes any process can start. This is usually called something like NPROC.

Don't put them in the shell script to begin with. Redirect them for example by reading a file or using an environment variable.

Are the columns mentioned in the WHERE clause indexed?

What database do you use for the XMETA database? Use a regular client for that particular database. Note that affecting XMETALOCKINFO directly is neither supported nor recommended by the vendor. You should use the Web Console wherever possible.

I have a vague memory that you need double semi-colons. (;;)

Probably. Based on your previous post, it's probably not to overload your system so much. But you may also like to verify the indexing in your local DataStage repository.

In short your system is overloaded. There may be an issue with starting processes, there may be issues with internal messaging and signal handling, and you're definitely running out of memory.

Lookup stage does lookups, not joins. You can emulate both inner and left outer join, depending on the Lookup Failed rule.

Merge stage definitely does not do a join; it follows a master and update(s) model. Again, you can emulate joins, this time via the Unmatched Master property.

You need to be aware that DataStage moves stuff (data) around in large chunks (3MB by default though this is configurable). To create heap files for a heap merge sort, the files themselves will necessarily be large. But they're only temporary, so don't worry about it. Just make sure you have PLENTY ...