DSXchange

Are you sure it's not DS_JOBS in the message?

I can't find any reference to DSR_JOBS (as a file) in any of the resource files.

Show us how these "variables" (which look suspiciously like input column references) "point to" environment variables.

That's good. Be very, very careful with the commands below, which are to be executed in your broken project. Get into dssh in that project. First, set a pointer to the VOC in the good project. SETFILE good_project_pathname/VOC GoodVOC Second, create a Select List of the DS verbs in the good VOC. Thi...

OK, assuming you did this in the project (and not in DSEngine) account, your VOC is incomplete. How badly incomplete it is impossible to tell. Do you have a project to which you can connect and which works properly? If so you may be able to get many of the missing routine entries from there. I await...

Can you get it to work using regular job parameters? Take the parameters file out of the picture for a moment, for the purposes of diagnosis.

Second, ought you not to be passing the environment variable value rather than the environment variable name?

mac4rfree wrote:The error is
Operator terminated abnormally: received signal SIGSEGV.

Are you sure this is a server job? Server job error messages do not usually mention operators in this contect.

indeed!

There are several ways, all documented in the DataStage Administrator manual.

The most common is for stopping DataStage, and for starting DataStage. The uv command exists in $DSHOME/bin directory.

There is a selection tab in the CFF stage. You can use this to pass just the detail records. Create other output links, or use separate CFF stages, if you actually need to process data from the header/trailer.

Once you get version 8 you will be able to do all this with a single CFF stage.

Show us precisely the call you are making, and describe where you are making it from. If an expression is involved, did you build this using the keyboard or the expression editor?

Yet

jherr22 wrote:I need to get the nextval and join it with every row of an input file

could be read the other way.

No, a sparse lookup will get the next value and the next value and the next value... To get just one next value you can use a regular lookup (probably with Entire partitioning). Provide a constant in an additional column against which you can effect a faux comparison. SELECT 1,sequence.NEXTVAL FROM ...

Maybe you don't have INSERT privilege to the relevant XMETA table, or maybe you are using a job name that is not valid. Unfortunately "they" have not given us access to the source or listing code for Edit serv.Impl.jobDefnImpl so we can't work out what's supposed to be happening at address 0x5ed0192...

With no idea what your SELECT is actually doing, no. Consulting with your DBA would be a good idea, particularly if your SELECT includes any of WHERE, GROUP BY, HAVING or ORDER BY clauses. There may also be a "fast export" alternative available.