DSXchange

It only fails on my workstation. The problem is that the download of ds did not complete or failed. DS installs, but only partially. It uninstalls, but when re-installing, uses the partially downloaded ds package. I believe if I could delete the download, then force a new download, all would be well...

conclusion:
the download of DS did not complete,
the install seemed to cope, but resulted in a partial install
the uninstall did not delete the download

thanks for your help

Runtime Column Propagation is not enabled and the job design does not have empty column grids. It seems that every Processing Stage has the column tab empty. Even transformers have empty column panels (the two lower ones). I have completely removed DS (even used regedit to totally remove), reinstall...

[quote="ray.wurlod"]Obvious question, but someone has to ask. Have you PUT any metadata onto the link (that is, have you loaded any table definition into the Columns grid on either end of the link)?

I have this problem even when I open an existing functioning job

After a new datastage 7.5.1A client installation, the column tab of processing stages have no content. It is a tried and prooved "network" datastage installation on a "supposedly" stock standard XP workstation designed for data stage development. Should I be looking at operating ...

All is well now,
Added copy stage between Transformer and Teradata API stage and cleared the partition in the copy stage.

Thank you for your help

How do I tell which stage the warning is coming from?
I'd assumed it came from the sort stage...

The output sequential file stage had the collector type set to Auto.
I changed it to Sort Merge and selected CATEGORYTYPEID as the key, compiled and re-ran the job, but it gave me the same warnings.

Hi, I have a sequential file going into a sort stage. There after it goes to a transformer, then onto another sequential file and dataset. The job seems to work ok, but it gives me the warnings: main_program: Sort key "CATEGORYTYPEID" no longer exists in dataset schema. It will be dropped ...

Hey hey, that was cool... i needed to alias the column name a."bad-name" as bad_name before it stopped grumbling. Maybe this is a version 8 thing?

anyway, thanks again for your help ray

It is a parallel job, but if need be, I could use a server job

that helps the select statement, but one still needs to define the output type. For example select a."cmp-id" as cmp_id from tbl_a a is a valid SQL statement, but in the outputs column section it still expects cmp-id, which can not be entered. I created a server job to do the same, and alt...

I need to select from an ODBC progress database, but all the column names have dashes instead of under scores in them.
Can anybody point me to the correct syntax for the select statement and technique.

sorry, but I couldn't find it with search

It all seems a bit buggy to me. When the task is created in a server job, and you click on reset from the director, it will execute (actualy, any user defined query seems to execute. For example, we tried a test "wait 15 minutes", and the reset did just that). Not to be perturbed, we re-wr...

We have a job that has a source odbc stage that calls a SQL Server stored proc, which inserts a record into a table and returns the same record to be processed. It works! But, when one resets the job via director, it also executes the stored proc. Any idea why the stored proc gets executed when the ...

Thank you both for your feedback.
I know the name of the orphan datasets as their raw data filenames contain the name of the user who created them.
I shall go through my four temp directories and rm away the orphans.

thanks again for the prompt advice.