DSXchange

Why do you believe that the Sequential File stage does not allow using Ctrl-Y as the field delimiter? How have you tried to enter that value into the Field Delimiter property?

Assuming all the source files have identical structure, use a Sequential File stage to read them using File Pattern, and have that stage generate the file name into an additional output column.

Once you know what it really is, a simple Convert() function to remove it - to convert it to "" - will do the trick. Imagine that it happens to be the ASCII NUL character (0x00). Declare a stage variable svNUL of type Char(1) and initialize it to Char(0). When tidying up your string in the...

I can't get my head around CHAR with no length. CHAR is defined as fixed width - that just doesn't work with unbounded. VARCHAR, yes...

So, interview questions aren't part of the "real world", Craig?

I would also add: communicating with IBM support (or your official support provider if that's not IBM) and maintaining local records of support requests (and responses), APARs, quick fixes, patches, etc., that apply to your site periodically reviewing Fix Central to determine whether there are any p...

Barely enough!

dsjob is not in VOC. Period. dsjob is executed from the operating system shell. Don't worry about VOC. It's just there. It's part of the engine. There's one in every project, and a couple of others besides. Don't try to open it with any operating system command; damaged VOC files are pretty much ir...

It is recommended practice to have UVTEMP pointing to somewhere other than /tmp (leave /tmp for the operating system). UVSPOOL is not used by DataStage; it is a hangover from the days when DataStage used UniVerse as its engine. But there would be no harm pointing it elsewhere, provided that that &qu...

You always truncate your messages... kkkk You can get yourself a premium membership of DSXchange that will allow you to read the entirety of posts from five posters designated as "premium posters". Premium membership is the main way that the hosting and bandwidth costs of DSXchange are me...

Welcome aboard. You don't use the ORCHESTRATE object in SELECT operations at all. This object is only used in target stages.

It doesn't. If you want that, you have to code for it. Create a "recovery" job. Incidentally, it's a sequence, not a sequencer.

Funnel puts all rows from all inputs onto its single output. That's a UNION ALL. There is no capacity within the Funnel stage to remove duplicates. Remember, one task, one stage type.

The DSN is localuv (which should be preconfigured in the uvodbc.config file). The UniVerse and UniData stages primarily exist to access those databases, both of which are sold by Rocket Software. The local DataStage repository remains reasonably compatible with UniVerse, but the UniVerse stage is no...

Instead of counting you could sum the constant 1. That would give you the options I mentioned.