DSXchange

Edited my previous post to correct typo.

Yes, provided you used the correct map. It may be that your "Asian characters" are not encoded using UTF-8. There are many other possible maps.

Yes, but the UNIX "black hole" is /dev/null, not NULL. In Windows it's .\NUL (there is one in every folder). Hmm, maybe that tells us something about the O/S! I still prefer the Copy stage. Add an annotation and/or a description in the Copy stage to alert the next developer to what you intend by hav...

Parallel jobs can be compiled in trace mode, in which case all column values on all links are traced using Peek events in the log (hint: don't run very many rows!). Active stages in server jobs can be traced via the Tracing tab on the Job Run Options dialog when submitting a job run request. There a...

Indirectly via an after-job subroutine. Better (in the sense of more easily maintained) would be via a Routine activity in the job sequence that invoked the parallel job.

dspxlearn wrote:What if you want ONLY the rejected records(unmatched) records from the lookup stage and the matched records are not be be captured anywhere?

Send the matched records into a Copy stage with no output link. It will eat them happily.

Don't you need percent signs on the time components in the format string also?

Search to learn the cause. Then take appropriate steps to prevent. Run dsrpcd. Don't run the script under different user IDs at too short an interval.

Write out the precise specification of what you want to do. You will, as Arnd identified, be reading rows from the files as single strings. Therefore you could use one of the comparison stage types for the column headings question, having dumped the table column names into a file. Why does it matter...

I always use a state file and Surrogate Key Generator (or SCD) stage. I guess that's why I never spotted it.

Welcome aboard. The answer is "whatever your particular DB2 installation requires". Consult with the DBA.

You got AN answer. Care to mark the thread as Resolved?

For something that simple, you could nest the Field functions.

The MatchField() function can also be used here.

Get your DBA to check whether there are any reported failures at the Oracle end, for example "failed to start listener" or "no more listeners".

Do you get row counts (even if zero) reported on all ten output links?