DSXchange

As noted, there was no solution posted in that other link. However, there is a debugging technique mentioned by Arnd, I would advise you to attempt the same.

An Execute Command stage could count the number of records in the file: wc -l <filename> After that you just need triggers from there to two Job Activity stages, one could check for $CommandOutput=1 and the other could be simply an Otherwise trigger for when it is greater than one. Assuming the resu...

You can't pass "parameters within a parameter", any more than Job Parameters can reference another parameter within them.

Hmmm... I think someone just called me a chicken.

As noted, talk to your UNIX Admins. Once they help you get it working from the command line outside of DataStage, you'll know what the exact syntax is you'll need for the Sequence job.

That second NullToEmpty() serves no purpose.

Worst case you can have a Sequence job run the batch/script if DataStage needs to be involved somehow.

There's no reason to think it won't.

Not sure why you would need to job for this. All it should take is the O/S command "move" on Windows or "mv" on UNIX or via the MKS Toolkit. Create a batch file or script to do what needs doing.

Your jobs can do whatever you can code them to do. If you want to be able to pick up where you left off, then you either need to keep track of how far you get or have some kind of post-process that can determine that for you. Then a constraint in the job can use @INROWNUM compared to a job parameter...

hsahay wrote:Attempt to convert String value "28 NOV 10" to Date type unsuccessful.

That's close but you need the dashes between the segments for it to work properly.

Keep the target as a DATE in the Oracle stage but make it a varchar in the job. How does the date come in from your source? Typically one would use IConv/OConv to convert it from whatever your source format is to the format your DML is looking for, from your example it would be looking for "DD-...

You need to explain what all those "not working" statements mean. What's not working about them? Are you getting errors? Also, are you ensuring that your date is in the format that matches the mask used in the generated DML?

Never mind the fact that, because of the nature of your data, the join is giving you a cartesian product.

No need. Sequences must abort to restart properly and that includes your Master so make sure the master aborts whenever a child sequence aborts. You should be able to use "OK" triggers to accomplish that along with the "Automatically handle" compilation option, that or explicit u...