DSXchange

Job sequences are server jobs. You can copy the generated Job Control code from a sequence and post in a regular server job and it will work in exactly the same way. Sequences are executed by the server engine (using DSD.RUN) in exactly the same way that server jobs are executed. Yes, I am sure. And...

This one is tricky to diagnose. The "?" characters suggest that you may have used non-ASCII (Chinese?) characters in a property that requires only ASCII characters. Are there any related warnings or errors in the log?

Until you disable operator combination you will have no idea which stage is throwing the error.

That's not what RCP does. However you can use DataStage to trace back. Right click on a downstream stage and ask DataStage to report where a column comes from. You can select an individual column to trace. Another approach might be to import the job into another project and perform a cross-project c...

Seems to me you're looking for the DateFromDaysSince() function to generate the new date. Presumably you have already worked out your strategy - a file maybe - for saving the new date so that it can be picked up in 15 days' time.

DCOUNT is a reserved word (it's the name of a function). Use a different name for the stage variable.

That one is an alert only, because you have a date coming in to the stage (I assume from the name that it's a pivot stage) bit the column to which you have it mapped is VarChar(10). The message is alerting you to that fact. Use consistent data types across the stage to prevent this error. If necessa...

One from left field: is this job designed as a multi-instance job?

As always you're making it hard for yourself (and us). Disable operator combination so you (and we) can discover which stage (operator) is actually throwing the error.

In server jobs and BASIC Transformer stages there is the concept of a "null string", represented by the system variable @NULL.STR. It would seem that you have these in your data. Have you tried trimming Char(128) from your string? That's the DataStage internal representation of null. Other...

I don't believe it's possible to get DataStage to perform a Cartesian product with a single stage.

You might, however, perform two Left Outer Joins with transposed left and right links, and combine their outputs with a Funnel stage.

Here's what I have in mind. SharedContainer +-----------------------------------+ | | | SeqFile Output | | +-----+ +----+ | | | | -----------> | | ---------> | +-----+ +----+ | | | +-----------------------------------+

Can this field be legally populated into the target?

Is any error/warning message thrown?

No. Ultimately DataStage is an ETL tool, not a reporting tool.

No, you will need a bit more complexity than that, perhaps a lookup against the target using the natural key.

Have you considered using a Slowly Changing Dimension stage?