DSXchange

Compiler "cl" not found.

is pretty clear. You need a c++ compiler in order to compile transform stages. This is documented clearly (and often) in the docs as well posted many times here on DSXChange.

You need to upload the image to the internet, perhaps using a free service such as www.photobucket.com, then point to that.

You haven't reached the inode limit on the directory, but I am not convinced that your user-id has write access to that directory. Are you certain you used the same userid?

Could you please post the result of "df -k ." when attached to the project directory? Also, did you try the "touch" with the same userid as you use in DS?

attach to your project and type "touch ICanWriteTheYesWord" - is the file created? If yes, type "ls -al | wc -l", how many objects? Also type "df -k ." and see how much free space is available.

Yes, but we don't know which character set the poster is converting to in the job - perhaps not all characters are mapped.

These things would be much easier if the actual errors were posted earlier.

The error message indicates that either
a) your file system is full.
b) your inodes are exhausted (too many files in the directory or mount point).
c) you don't have write access to that directory.

I can't check now, but I assume those aren't mapped in the poster'S Windows 1252 - {whatever set they are using}

In that case you might have a corrupted account, particularly if you cannot execute "save as". If you can ensure that no other users are logged into DataStage, you can execute "DS.CHECKER" from the TCL command line to determine if you have problems.

No, unfortunately not. I checked up on the Server NLS functionality as well, since there is support for Japanese and Chinese calendars but as those are based 1:1 on the Gregorian they are easier than the Hijri calendar. There are a number of public-domain c++ routines floating around the internet, s...

Are you sure you used the same userid - or could you temporarily chmod that file to 777 to test? Also, is it just the one userid or one job that give you the compile error?

I would use the sequential file filter and use "tr -d '\201\215\217\220\235'"

Is your DB2 Database partitioned - I think that might be the determining factor. Or have you specified "sequential" for the DB2 Enterprise stage?

Do you have RW permission as your user to the DSParams file in your project directory?

Are you logged in at UNIX with the same userid as you used to compile the job? If you create a dummy job (Column Generator -> copy stage) and compile that, can you run that one from the command line?