DSXchange

If you're normalizing on the multi-valued field, you don't need an exploded sort. I wouldn't be a problem with UniVerse, but I'm not sure with UniData. Try BY.DSND rather than BY.EXP.DSND.

Red Brick Warehouse is different. In this case, the driver manager searches for $RB_CONFIG/odbc.ini.

If you think about it, there's no valid "transformation" you would do with these data types.
DataStage can move them, but only if they're encoded into a text stream first. But why would you use DataStage just to move things, when FTP is so much cheaper and, some would argue, simpler?

Does Oracle still use the word "easy" in the SQL*NET configuration tool? I always felt this was false advertising.

Make a hashed file with one row and two columns. The key value is a constant, say "X" (or something meaningful, like "SelectionDate"). The hashed file is loaded from the text file with the YYYYMM value as the non-key column. You can use this in a reference lookup within a Transformer stage. Or you c...

Not really. If you import from a DCLGen and the foreign key relationships are stated in that, then DataStage will detect them. You can manually edit DataStage's "table definitions" that result from the import, and thus identify foreign key relationships, but I don't believe there is any way for Data...

After changing SYNCALOC, did you run uvregen and re-start DataStage? If not, your SYNCALOC change will not have taken effect.
Have you adjusted the SPINTRIES and/or SPINSLEEP parameters?

Display length can be different from character length in character sets like Chinese, where there are double width characters. Thus, for example, it needs ten display positions to display five double-width characters, even though the "character length" of the string is 5. And, indeed, there's a thir...

What happens if you try

You probably have a mismatched format. The only time when strings containing dates can be reliably compared is when they comply with ISO 8601 (for example YYYY-MM-DD). What is the source (text file, database table)? What type is the parameter? Date type parameters are converted into DataStage intern...

No. Mainframe jobs generate COBOL code and JCL to get it compiled. These are uploaded to the mainframe where they are compiled and run. Whether any parallelism occurs is not under your control. Parallel jobs generate code that can be executed as osh in the Parallel Extender environment on the UNIX m...

No, it seems to be trying to read record "21849" : @VM : "1" - the delimiter is what causes the 1 to appear on a new line in the logged event. Somehow you're not properly normalizing the data. BY.EXP.DSND is not part of the WITH phrase - it is a separate clause. To use an SQL analogy WITH maps to WH...

Dunno. The "approved" method would be to do the two reference lookups in two consecutive Transformer stages, passing the results of the first along the link between the first and the second.

The Lookup stage is described, with detailed examples, in Chapter 19 of the DataStage Mainframe Job Developer's Guide (mainframe.pdf), which is in your DataStage manual set (in the Docs folder with your DataStage client software). They are also covered in the DataStage Essentials (DS390) class run b...

This is Universe not Oracle. Very funny. You probably wrote your own explain plan for UV, right? I just wanted to show how to bring in the job name. UniVerse (and therefore DataStage) also has EXPLAIN. It was introduced with the SQL Engine in version 7.3, and heavily refined in version 9.4. SELECT ...