DSXchange

Please advise what these issues were, and how you discovered and resolved them. This will assist any future searchers who have the same problem. Then please mark this thread as Resolved using the green button at the top of the page.

It might be better to use BY.EXP.DSND rather than BY.DSND.

Redirect the output of the uvsh.exe command into a file as required.

And what do you mean by "global"? That is, what scope do you see these variables having - for example within a sequence, within a project or across multiple projects?

Could be worth checking the XMETA database to determine whether the job name property for an IA job is data type VarChar(64). At the moment (weekend) I can't perform that check. But that would be my guess as to the cause.

This is a classic fork join design as others have indicated. Downstream of the Join stage use a Filter stage or Transformer stage to allow past only those records that have a count of 1.

HashedFile | V ---> Transformer ----> | V HashedFile The incremented value is written back to the hashed file. Neither read cache nor write cache is used. The hashed file contains a single record whose key value is a constant, for example "X". Of course you could do the whole thing with s...

On your server look in RT_BPxxx subdirectory, where xxx is the job number for your job. The file JOB.1140653906.DT.1562646482.TRANS3 contains source code created by compiling a Transformer stage. Look at and around line number 181 to determine what was happening there. A non-numeric value has been s...

I liked "Which of the Marx Brothers could not speak?" (it came out of deep left field).

The correct answer is "none of them". Harpo could speak, he just didn't in the films.

Use a reject link to capture any rows that could not be read. Do they have embedded quotes ("Subject referred to "The Three Stooges" for remedial therapy.")?

Who would that be?

Did you export this job from a system where NLS is not enabled to a system where NLS is enabled?

The "End:" is not correct syntax (it should not be there at all), but I would expect that that error would have been found at compile time. It appears that a sed command finished successfully (it was checkpointed) and whatever is next in the sequence has caused the sequence to abort. What ...

There's no default. There are separate import mechanisms - the Connector wizard or the Oracle import.

A more technically accurate solution is probably to set the quote character to none.