DSXchange

It's not in the Manager or in the VOC file for 7.5x2. What is the date/time modified on theirs, and what is the date/time modified for one of the "supplied" routines? In passing, I do note that there are two new supplied before/after subroutines (well, I haven't explored in there for some while) cal...

The graphical user interface looks just the same, but there's a UV stage on the start of the dashed arrow. Within the Transformer stage, however, you will now find a check box to allow multiple rows to be returned. Search on-line help or this forum to learn more about multiple rows being returned fr...

How are you trying to "use" Select List #0 ? SELECT VOC WITH @ID LIKE ...TMP... At this point Select List #0 is "active", the prompt changes to >> to indicate this. The DELETE.FILE command will use an active Select List, so all you need is DELETE.FILE Beware that it doesn't ask "are you sure", so be...

The ideal stage type is a QualityStage stage that invokes a QualityStage STAN (= standardization) job using the appropriate name rule set (e.g. USNAME for USA, AUNAME for Australia, GBNAME for Great Britain, and so on). This will bucket the name into title, name prefix, main name, given names, name ...

Why aren't you using a Hashed File stage? If you're doing everything in code then you've wasted the rather large cost of DataStage.

The second argument of DSRunJob specifies the run mode; normal, validate or reset. The result is always retrieved using DSGetJobInfo; in this case with a second argument of DSJ.JOBSTATUS. For the current job you can also use the DSJobStatus macro. A validation run will result in a status of one of D...

Open the Hashed File stage (note that it's hashed file, not hash file). The name of the hashed file is a property of the input or output link.

A much better solution is to check for this before processing the file, not least because you will end up doing a lot less processing. Use the UNIX command tail -1 filename to report the last line of the file, and use something like grep ^999 to determine whether the line begins with 999. Pipe the o...

There's no stage for performing a vertical pivot (which is what you are asking for). Search the forum for "vertical pivot"; you should get at least two techniques, each involving an intermediate passive stage.

If you want to make it pellucid, specify DSGetParamInfo() explicitly. InputFileName = DSGetParamInfo(DSJ.ME, "input_file", DSJ.PARAMVALUE) OpenSeq InputFileName To FileVariable... You should always specify a LOCKED clause with OpenSeq. Since OpenSeq sets an update lock, you need to...

Do NOT use read cache or write cache, and ensure that every read from the hashed file has "lock for update" set. Under these circumstances you will not experience any concurrency issues.

kumar_s, would you care to share your solution, in particular with proof that it handles leap years correctly (2000 is a leap year, but 1900 and 2100 are not)? Remember, too, that this is the forum for parallel jobs, so an expression suitable for a parallel Transformer stage is required.

Add an extra Transformer stage just ahead of the ODBC stage. Create a reject output from this Transformer stage. Rows will go down this link if they are rejected by SQL Server. In addition to the original data rows you can create extra diagnosic columns on the reject link. The can be populated with ...

I vaguely recall that there's an option in the ODBC driver configuration that allows SQL_SUCCESS_WITH_INFO messages to be suppressed.
Try searching on the forum or ADN, can't recall whether this has been posted before.

The boredom can be relieved somewhat by creating a job template with all the stage types, links and job parameters already in place. All that's required is to import then load the table definitions.