DSXchange

Don't bother, is my advice. The DB2 Enterprise stage should automatically (if you've told it the partition table name) use the inbuilt DB2 partitioning.

Performance tip: Initialize a stage variable (let's call it svTab) to Char(9). Do not provide a per-row derivation for this stage variable. Use svTab rather than Char(9) in your expressions. This way the Char(9) function is evaluated only once for the job, rather than once per row. Char() and Seq() ...

Tell you what, you give us your take on what each of these does, and we'll address your impressions.

Include DSParams and .*.adm from each project in your list of files.

Best practice is to test everything.

Time and resources sometimes mitigate otherwise, so you need to identify the most critical components and test those.

No. Almost everything is DataStage is based on characters, not bytes.

You can, of course, manually edit the table definitions, even update the entries in the repository if you can find them, but it's probably easier to re-import.

Welcome aboard. To add to Arnd's suggestion, may I add the idea that you initialize a stage variable to Char(10)? That way the function need only be executed once, rather than once for every row. The stage variable (let's call it svLF), having been initialized, has no per-row derivation. The column ...

My take on the term is circular reference, for example where variable svOne depends on variable svFour for its definition while at the same time variable svFour depends on variable svOne for its definition. In practice this ought not to present a problem provided both have been initialized because, ...

Have you checked that the keys for rsh have properly been emplaced?

Parse the output to determine the status of the job and whether a reset is necessary. If it is:

If you are reading the file in sequential mode, then you can skip the first 12 million rows (or whatever the number is) with impunity. If you are reading the file in parallel mode (perhaps more than one reader per node), then it's still possible, but you need to check how many rows have been process...

Based on your rather loose description it certainly looks like the best - most flexible - solution is a lookup table of some kind. You could do it in a routine but, being a parallel job, the routine would have to be in C++ and, therefore, would not have the Replace() function (unless you created tha...

I had a doubt because version 7.5 had the limitation of 20 stage variables. There has never been any limit on the number of stage variables (not counting very early versions of DataStage in which there were no stage variables at all). Perhaps you are confusing this with the number of intermediate r...

Can you mark this thread as Resolved now?

And that will cost you a bomb in extra licensing charges.
Use a node pool and run the BASIC Transformer in that - you can use four nodes from the machine where the server engine is installed and at least get some degree of parallelism.