DSXchange

I have done the same types of deployment with overwriting objects and have never seen this error message/warning. While it doesn't look like the message is affecting you, I would submit the problem to IBM via your support provider.

While BASIC transform stages have ASCII and EBCDIC functions the better way to solve this splitting would be to edit your metadata or COBOL definitions to split the data before reading it, this would be a COBOL "REDEFINES".

You can safely do a RESIZE of the DS_TEMPnnn, RT_CONFIGnnn and RT_STATUSnnn to static hashed files with a modulo of 1, this saves 3 subdirectories and 9 inodes per job.

Column export will do it as well.

Normally there is no need to explicitly convert in a transform stage. What are you doing to the value in teh transform that generates this message? What is your data source and what character set is it in?

The DS_AUDIT hashed file contains this information. This thread will get you on your way to using the DS_AUDIT hashed file.

One can do it in PX, but not as easily. Take a look at this thread

I would guess that the DataStage user has only "R" and not "RW" rights to "tmp//meas.ds" because of the error message

"Unable to open descriptor file to update: Permission denied "

If you keep the same character set throughout there will be no problem with such characters. 1. What is the encoding used in your DB2 table? Have you told that to DataStage? 2. What is the encoding used in your Oracle table? Have you specified this in DataStage as well? When you fix #1 and #2 your d...

Debug it by checking the value of that string at each stage of the process to discover which component seems to be doing the modification.

I just tried using the search functionality and found 16 matches on the string "Input dataset 0 has a partitioning method other than entire specified; disabling memory sharing". Were none of those any help to you?

The first link I looked at had both a solution and explanation.

Does "View Data" function on both Oracle stages?

The routine will read the first line of the file specified in "FileName" and return that string.

There are numerous ways to do this. If you also have packed decimal data then the cmoplex flat file stage is best, otherwise just use a sequential file stage and specify that the data is EBCDIC in the format tab.

karry450 wrote:... DSP.Open at address 574...

What kind of stages are you opening?