DSXchange

Convert all the valid characters to "" using a single Convert() function then test using Len() function to determine whether anything remains. The following expression is true if TheField contains a valid value, false if not. Len(Convert("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ&.,"...

OK, we're not getting enough from the error message alone. Somewhere, perhaps in your source data, there is a value "R" that is not being successfully converted to a timestamp. Can you please therefore post (a) your job design and (b) the column definitions of your source data?

Initial skepticism is what I've always found to be best. Disbelieve everything you are assured about data quality (if it's good) and CHECK. Profile the data to learn what's really there (ProfileStage is good here, and AuditStage is good for verifying compliance with business rules*). Based on the pr...

Alas, that is not how the product works, unless you create your own stage (probably a Build stage). In the "out of the box" stage types, you configure the processing requirements of each field separately.

Ah, the Force option in the Copy stage itself. Well, that's basically there so that you can require the compiler to include a copy operator even though one is not, strictly speaking, needed. Another way is to use the Copy stage to do something, even something trivial like re-naming a column. You say...

This is not a hashed file issue. It is an issue with your Oracle query. The ORAnnnn form of the error message code indicates that this is so. There are several possible causes of "snapshot too old": enlist your DBA to help to track down why yours is occurring. The message is being generated by an "i...

Search the forum for "BCI" (BASIC SQL Client Interface) for one approach, however this will need an ODBC driver. You might also search for techniques using sqlplus invoked from a shell script.

Original edited.

Don't know where you read that. The Copy stage is compiled out irrespective if it's not needed (that is, makes an identical copy of its input). Force compile has nothing whatsoever to do with it.

Search for CENTURYPIVOT both here and in the manuals.

Because you are relatively new, let me gently suggest that Search is your friend here. Yours is a fairly common requirement in DataStage and others before you have posed exactly the same question - and had it answered. Yes you can use Upsert, but this tends to be less efficient; for example "update ...

Open the file with WordPad rather than Notepad. The squares are newline characters: AT uses UNIX line terminators. You can, therefore, use a Sequential File stage that specifies UNIX-style line termination. Otherwise, the WordPad output will give you all you need to know about parsing the output. Th...

It just occurred to me that, since you apparently don't want to use a job sequence (not a Sequencer - that's something you have IN a job sequence) that you could also effect whatever tests you like in a before-job subroutine. Setting the ErrorCode argument to any non-zero value will cause the job to...

The only way you will do this directly from UNIX is if you have a UNIX-based ODBC driver for MS Access. No such driver ships with DataStage, so it is up to you to find one. An alternative is to dump your table(s) from Access into some kind of flat file format on Windows, FTP those files to UNIX, and...