DSXchange

Umm... don't use vi ?

Examine the file with a hex-capable editor to determine what the line terminator character(s) (if any) are.

What precisely do you mean by "corrupted"? That is, what error do you get? What raw record is sent to the reject link?

Declare the output column to have data type dfloat.

Limit the report to those events whose timestamp is later than the timestamp of the "job started" event for the current job.

Welcome aboard. :D

These are fairly well described in the Sort stage chapter in the Parallel Job Developer's Guide as well as in the on-line help for the stage. Have you looked at either of these?

Please post the results of your research. This is a sharing site.

There may be some corruption in DS_JOBOBJECTS or DS_JOBS, or they may need to be re-indexed. Because you can count DS_JOBOBJECTS I will assume for now that that hashed file is not corrupted. Can you count DS_JOBS? At a time when no-one is using DataStage, login with administrator privileges and exec...

Have you considered using a Complex Flat File stage, which has the built-in capacity for dealing with multiple record formats?

Or pre-processing your file, adding sufficient delimiter characters so that every record format has the same number of columns?

Not a DataStage problem.

Does the new member sign-up process include information about, and the ability to sign up for, premium membership?

Welcome aboard. :D A CSV can have as many lines (rows) as you like - the limitation is in some of the tools that use it. Use your DataStage job to write to a single file. Then write another DataStage job, or a BAT file, to split that file into pieces each containing no more rows than your tool can m...

Define "performance".

Why?

If it's working properly in DataStage, leave it alone.

The "problem" seems to be in your data viewer at UNIX level, not in DataStage.

My thought is that you have not specified exactly how you want the output to appear. If you have, then a Fmt() function wrapped around two Field() functions would do the job. Or multi-concatenation of fixed strings and the results of two Field() functions. There are other solutions too.

Welcome aboard. :D

When importing are you referring to a valid DataStage export file? If you exported in XML format, for example, you will need to choose the XML import option.

The \ in the od -c output leads into a representation of a non-printing character, such as \09 for tab. The fact that View Data sees it correctly suggests that it is fixed-width; the fact that your UNIX view does not suggests that the viewer is expanding tabs.