DSXchange

Not talking about your source, but I see I misread your post anyway. You say you need to produce a "fixed width file" so we need to know what properties you have set in the target sequential file stage.

Fixed-width files don't have field delimiters, what happens if you remove that property all together?

Betty, please start your own post rather than jump on someone else's. That way you can include all of your information including the error(s) you are seeing, even if they are "the same problem" as this poster's. Also make sure you let us know what flavor of UNIX you are running.

Right, the default warning limit is client specific, so can change from user to user and is not something associated or stored with a job. You really need to supply it at run time, either using -warn n option from the command line with 'dsjob' or taking one second to click on the Limits tab in the r...

Perhaps not, as that's not really what those "project level" parameters are meant to be used for. They are meant to parametrize values that would change from project to project or environment to environment but would otherwise be 'static', not values that would change from run to run as yo...

OK, then what runs the Sequence job? Whatever warning limit that Sequence job runs with will be passed to all 'child' jobs that it runs.

How/where are you running your jobs?

But do you know the maximum number you could get?

Your method is fine. Another way is to write them to three separate files and then concatenate them together after job in the proper order.

Nice question marks.

Never seen it before, no. What does the Director's Job Monitor show for those links?

As you noted, those BLOB and LONG data types are not supported so your best bet is to stay with an Oracle solution, probably PL/SQL.

ArndW wrote:assuming you created a normal hashed file instead of one with a path.

Hey, everywhere I've been the pathed hashed files have been the normal ones.

"orachadmin"? You misspelled the command.

sreddy wrote:read the data in the job1 out put dataset perameteraized that value and pass on to append job2 out filename.

Can you... rephrase... that? It doesn't make much sense at the moment.

As noted, Field() will help, but could get a little bit tricky if there are more than two 'words' in the string. Perhaps pull the first field as first name and everything else as the last name in that case.