DSXchange

abc123 wrote:Setting APT_ORAUPSERT_COMMIT_ROW_INTERVAL to 1 would make insertion/updation extremely slow.

Curiously this is not always the case.

Are the character strings quoted?

Use a constraint experession using Convert() function to detect the existence of any other characters. Something like the following would suffice:

If you click in the "blank" area of the Connector stage, does the row of the grid on which you clicked become visible? Can you then click elsewhere to have other rows become visible? And are you connecting through some kind of tunnelling protocol, such as Citrix MetaFrame?

Problem with shared library search path not been set up correctly? Or, perhaps, permissions on that object or one of the directories in its path?

Use Wordpad rather than Notepad. Wordpad seems to handle both kinds of line terminator gracefully.

I am running 8.5 on AIX. There is no Command stage. A search revealed only four objects with Command in their name:
Execute Command activity (for sequences).
A folder in the repository Data Elements branch.
Two data elements COMMAND.OUTPUT and COMMAND.RTNCODE.

No Command stage.

Look on the Columns tab. How many columns are mentioned there?

Have you used the "data rules" profiling capability (ex AuditStage) in IA? Discovery doesn't have - and is never likely to have - these, though a tool that could discover data rules would be really magic.

Its file in the &PH& directory is open while the job is open. You can't delete a file that is open for writing. Hence &PH& is not cleared. You would be better trying a command that deletes only closed files - maybe files more than a week old (as appropriate to your site).

You must have a DSN. Not a User DSN.

Therefore you must have an ODBC driver.

ETLstats works with version 8.5. What are you saying?

You have exactly the same issues on 7.5. Absence of the //PURGE.SETTINGS control record flags use of the project defaults.

Change the system clock on the server while the job is running.

Rows/sec is an almost totally meaningless metric.

I suspect (the metadata in) your job design only names two columns.