DSXchange

First explain why you think you need to 'escape' anything.

First off, there's no 'View Data' option in the stage. What do you mean when you say that you are "not able to see the records" in the XML file? Is it empty? Unless you are outputting 'formated' XML then it will be one long record, essentially. You'll need to transfer it over to your PC and use an X...

A good place to start would be to search the forums for "DaysFromDateSince", that should get you started. Branch out from there if needed.

Have your DBA help, check to see what's happening in the database while the job runs. 'Hanging' generally equates to having locking issues. Or it could just be very slow if you don't have an index to support the update key columns.

Well, what do you know, there is - all 658 pages of it:

http://www.redbooks.ibm.com/abstracts/sg247576.html

Free as a pdf or $60 printed. Hah, and only one comment there so far, not really suprising whom it is from.

And let's not forget Kimball's Data Warehouse ETL Toolkit. Not specific to any tool but chock full o' goodness.

Basically, because it generates metadata more closely typed to the source, something much more important in PX than Server. Plug-in metadata is notoriously... sloppy... at times.

No, not all. RT_LOG34 is the log for job number 34, the one you are having problems with per your error message. It doesn't go by name but rather by the internal number assigned in the repository.

Since you're only passing one row through, then that row equals "the end" does it not? Otherwise, consider doing it "after job", which means a routine or a script.

While corrupted hashed files can be 'repaired' it's not really worth the effort for a job log unless there's vital information in there. So, to clear it you need to run CLEAR.FILE RT_LOG34 from the Administrator or TCL prompt. Remember to reestablish your auto purge settings for the job afterwards a...

Rather than pulling teeth, why not lay out all that you needed to do from the beginning? Yes, link everything together and the row should flow through each stage.

I hope so, seeing as I'd already spelled that out.

1. Automatic.
2. Automatic.

Excellent. :D

Worth noting, however, that the files need to be sorted for the join to work. From the man page:

file1 and file2 must be sorted in increasing collating sequence (see Environment Variables below) on the fields on which they are to be joined; normally the first in each line.

Because you did not add a stage variable as I specified. No need to use it but it needs to exist.

Swell. Then that means it can support an output link as well as an input link and you can use the Forward row data option to pass the same columns out as came in.

No not directly, but you can pop a 'DataStage shell' using dssh and then issue the DS.REINDEX command from the TCL prompt.