DSXchange

Do an exact search here for "has keyprep export format error" and check the other two posts that come up besides yours and the one previously linked to. See if anything there helps. If it doesn't then it is time to involve your official support provider if you haven't already.

By 'reverse the mapping to a masked value' Ray was asking if you had a need to be able to undo or 'unmask' the masked value back to its original value. Hopefully not as masking is a one-way process. Not seeing the problem with SSN, as noted earlier all it seeems you need to do from the examples you ...

I would suggest you search here for DS.TOOLS which is another mechanism find PIDs and handle all aspects of lock management in your version. You can run this from the Administrator client.

As noted by Ernie, the difference is in your files. So, what is different between them?

Those are all about your metadata in the CFF stage not matching the file you are processing. 1. Your metadata shows a record length of 199 but the record ended at 196. 2. The field mentioned was larger than you metadata allowed. 3. You are processing multiple record types and 'PDD' is not a type you...

How similar?

Just a quick reply - '0000-00-00 00:00:00' is not a valid timestamp because '0000-00-00' is not a valid date.

Those instructions are for purging logs from the XMETA repository. For the hashed files I'd look into a looping script, probably simplest to find all of the RT_LOG names you've noted and issue a CLEAR.FILE on each one.

You don't need to escape them, use single quotes around the string with the doubles inside:
(and vice-versa works as well)

Dates have no formats, only strings do. Use DateToString() with a format mask that matches your desired output.

Are both fields strings?

So... one thing clarified, you want to delete records in spite of putting 'data loading' in the subject. Everything else is still vague though, as noted.

Five and a half years later and several things discussed... you have WHAT issue, exactly?

There is no global solution. DataStage server jobs inherently do not have data types. Context determines how comparisons are done (numeric or string, right- or left-justified). If both operands are numeric, a numeric comparison is performed. Interesting... I don't recall running afoul of that littl...

You need to learn the Parallel date handling functions, of which there are many. Primarily look into StringToDate() and DateToString() but a quick check of "Appendix B. Parallel Transform functions" in the Parallel Job Developer's Guide should be high on your list as well.