DSXchange

After you've found the compiler, just run 'aCC -V'

I intentionally use that command, since if it isn't in the path then you have other problems in DataStage.

Instead of the uncompress stage, use an external source stage and put the gunzip/uncompress command there. I've use that extensively with no problems.

Try to get the location

I seem to recall that you had to mount the CD with specific options on Solaris, otherwise it wouldn't process executables correctly. Can you check your installation guide (which will have the correct mount options listed) and ensure that you have done that correctly?

Looking up Abstract Syntax Notatation Wiki was interesting, but shows that there are numerous standard encodings possible and as far as I know none of them are 'in-the-box' in DataStage. As all versions of DataStage allow you to program complex code (be it in DS/BASIC or C++), you can implement your...

The status is not updated to reflect the state of a job if it is killed from UNIX (something that should only be done under extreme duress)

saikrishna wrote:...forceful stopping of the job in director and also in backend unix box...

There is no builtin CRC32 or similar routine in PX. But it is easy to link in your own C++ routines, and there are numerous public domain sources for all manner of checksum routines available on the WWW

How complex is your file? What format is it (does it have terminators)? You can use the split command to split the file into subfiles. If you write a job which just reads the CFF and doesn't output anything what speed does it read at?

Can you have your dbadmin monitor the process while you are running it - he/she will be able to see exactly which lock contention you have.

At least it is now a relatively easy thing for support to test and confirm. Please keep the thread updated with the result so that it is useful for future users of search.

The designer canvase might limit you to 128 or 256 or so, but if you use the old method with the Job Control inserted there is effectively no limit. You machine might get somewhat bogged down with that many jobs if they run concurrently, though.

Sounds like the perfect job for me. I can delegate just about anything, including responsibility.

where do you want to retrieve it? From a Job Sequencer? From a function in a Transform stage of a job?

How large are the DATA.30 and OVER.30 files when the error occurs? Anywhere close to 2Gb?