DSXchange

Sorry if this is a silly question, but is "Transfer" the name of the job or a stage / link name within the job?

In other words, the columns can be empty but not null. Null means 'Unknown' and is a database concept not applicable to sequential files.

Two problems. 1. It is not recognized as a valid function name, so it assumes it is an array that you've never dimensioned. If we were talking about a custom routine, I was prepared to tell you to include the "include" file statement you are missing. 2. The reason it is not being recognized is that ...

Was really hoping for a question to respond to. And actually, the informational message is because there is no where clause and it deleted every record from the table.

Hopefully, that's what they intended.

There you go.

Was going to suggest using "rownum" and parameterizing the "500" part in case it needed to change.

You should be able to handle this using normal (non-user-defined) sql and parameterizing the 'n' for the 'Top N'. Perhaps you can post your sql for our perusal?

Since it's looking for a relative path, suggest you cd to "/root/ascential" and then execute "./install.sh" instead.

Please be specific about what directory you are initiating the installation from and the specific command you used.

Fun! :D About the only thing I can think of on the DataStage side would be to enable tracing. Switch to the "Tracing" tab in the Run dialog and turn on at least the Property Values and Subroutine Calls trace levels. Maybe something will come of that, but I suspect that any "tracing" will need to hap...

dec1177 wrote:I had to include the path in quotes because of the fact there's a space in the pathname (I think).

Yup. Thank you Windows "long filenames".

Or you could have used the original recipe 8.3 "short" name without quotes, which is (more than likely) "Progra~1".

Bad Idea. Use the GUI, Luke! Create a new category and move everything over to it, then delete the old (empty) one. After you undo this, of course.

Unfortunately, no, not really. I mostly just wanted to verify that it was 'being executed'. I've no real experience with SQL Server so the gory details of differences between the 2000 and 2005 editions are unknown to me and thus I've got no clue why it would work in the older version and not work (o...

sajarman wrote:It seems to me that hashed files can act as good remove duplicates

Yes, very true. :D

Copy the cdrom to a directory on the Linux server and install it from there.

Your test1.ksh script doesn't have execute permissions for the user attempting to run it.