DSXchange

It depends on exactly how "flat" the file is. A Complex Flat File stage may be the more appropriate choice. If you want to form the join between two text file, a Merge stage is your choice. But, as Craig asseverates, a Sequential File stage is the most likely choice.

The timestamp in quotes is a format string. You do not require this if your timestamp matches the default timestamp format for the job.

To be precise, it is not 75% of 79 or you can miss 19 out of 79. The questions are weighted by category, I believe. So it is 75.949367088607594936708860759494% No, becasue you have introduced a statistical abnormality. 75% presumes 100 sample points. 75.0% presumes 1000 sample points. A percentage ...

You should be able to lookup on more than one column. However, you may have to "lie" that they are all "Key" columns on the link connecting to the reference input. That is, in your case, declare both columns to be key columns.

You do have it. Use the Expression Editor in a Transformer stage. It will appear in among the available Functions.

Welcome aboard. :D

You need to specify precision large enough to encompass all the digits in your number - at least (scale + 1).

When you imported the Oracle table definition, what were the precision and scale? What is the Oracle data type of the column in question?

Use bin/dssh rather than bin/uv

You won't get better help than that. It's an accurate and precise answer.

When reading the file, use sort -u as a filter command in the Sequential File stage.

Or, if you want a really slow solution, use the ODBC driver for text files, and use your SQL statement from an ODBC stage.

I make some allowances for brand-new posters. :D

But not in the case of misplaced or misused apostrophes!

Hi, Actually what i want is the DS internal table/report which stores the number of rows updated by a job and in which table. Which i am going to access by a unix script and inserting in a table which will be used for job contraoling. So i dont want to manually create report ofter each run. DS_JOBO...

You will need, at the very least, that the columns in the DB2 tables are defined as NCHAR or NVARCHAR rather than CHAR or VARCHAR.

T30FILE requires approx 112 bytes of memory per slot. SCRMIN, SCRMAX and SCRSIZE affect the size of scratch buffers. OPTMEM is the amount of space used by the query optimizer when some form of SELECT operation is performed (mainly by DataStage clients). JOINBUF is the size of the memory buffer used ...

An Exception Handler does not cause an abort.

A Terminator activity sends a stop request to all running job, so that their status will end up being "Stopped" rather than "Aborted".

Thus neither suggestion is correct. Neither approach will abort a job.

Welcome aboard. :D If you are writing your own code, involve the DSLogFatal() function. From a Job Sequence use a Routine activity that calls UtilityAbortToLog(). Each of these has a side effect of aborting the job. I disagree completely with the practice of aborting jobs, but the information above ...