DSXchange

We have also found the creation of the data load file very slow using the DataStage IQ plug-in and have resorted to creating the file using a sequential file stage. We let DS create the LOAD SQL.
This works in most cases except for timestamp/datetime columns. For these, DS LOAD SQL does not allow for delimiters (by default |).

So how does one create such a seq file or should I just create my own LOAD SQL, run the load manually and forget using DataStage for loading

Welcome aboard! :D

I do prefer to create my own control files, because I can then tune every little thing that I can to optimise the load. I do sometimes use the bulk load stage to generate a prototype; other times I don't.

If I want to run the bulk loader under DataStage control, I use either ExecSH as an after-job subroutine, or DSExecute from within a custom after-job subroutine.

Bulk loaders generally work best with fixed-width format data; this may be why there's no option to have delimited fields in the data file. If you think this is a deficiency, you can lodge an enhancement request here (there's a dedicated forum for enhancement requests to the products), or on Ascential Developer Net, or through your support provider.