DSXchange

Search is available in Director. Having found the job, select it then choose Tools > Run Designer. The selected job will be opened.

Manage expectations.

You can, of course, use a sequential file for lookups, by identifying the lookup key columns in the metadata and employing an ODBC stage via the ODBC driver for text files.

Just don't expect any speed.

Not to use DataStage at all. Use two wc -l commands wrapped in expr commands. Capture the results using backticks, and add within the scope of an outer expr command. Use output redirection to write to a target file.

OK, so the amount of time spent in the pre-loop code to build the lookup structure needs to be assessed. Also out some timing points into the per-record code to identify any hot spots.

This question has also been asked there.

Etiquette Note
Asking the same question in two different forums may result in your receiving conflicting advice; a server job answer may not be relevant to the parallel job environment.

Complex Flat File stage

1) diff (integer), source1row (subrecord), source2row (subrecord)

2) columns collection is a property of the link - it can't be different at the other end

3) depends on how you've configured the stage, but probably something similar to the original source input

Hire a consultant with DataStage and Oracle experience. Get the DBA to clean up the consequences of your killing processes. Get the DataStage Administrator to clean up the consequences of your killing processes. Cease relying upon killing processes.

Your job will fail if it tries to write to anything inaccessible.

NFS is totally contraindicated for scratch disk, and generally contraindicated for disk resource also.

None, apart from entering an enhancement request.

How did you create the project? As a general rule you can create projects anywhere except the root of a file system, though they should be on local disks. I suspect the problem opening RT_STATUS11 is either permissions related or because your T30FILE table is full, and unrelated to the location of t...

Is it a sparse lookup or a normal lookup? If the latter, a lot of the time might be being taken in loading the virtual Data Set from which the lookup is being performed.

Which "uv file" are you talking about? The uv file in bin should not be affected by the regen process.

Is this the same question as the one in this post?