DSXchange

Yes, there's a VOC entry you need along with the files and the test would have created that, although I would have thought you would have seen a different error. Just as an FYI, a search on that file's name could have turned up previous conversations where the advice is to test first to create the n...

Did anything with that name exist there before you copied it? For #1 can you cofirm it is a file and not a directory? Also (as a general statement) ensure the ownership and permissions are correct.

So... out of curiousity, how did you resolve this?

Ah yes, SDKSequences. Did you copy that hashed file over to the new project? A simple 'test' of the existing sdk routine would have created a new one for you if it is OK for it to 'start over'. Best to explain whatever it was you did so we can help figure out what it doesn't like.

Welcome.

Pardon my ignorance but how exactly does one 'test' the Connector, is that an actual option in the stage? I find that Oracle error to be odd as it is one thrown by PL/SQL rather than normal SQL type stuff.

Involve support.

What other FS exists for AIX? As noted, that's a question for your SA. And I would bet that "... and they are willing to support" gets added to your question. :wink: Our customer, for internal reason, doesn't want to divide project...it is so. Sometimes it's not about what is wanted but w...

Then it seems to me you have two options. One is to ask the vendor to make the change. Second is to pre-process the file and change it yourself using something like awk / sed / perl. This could built in as a Before Job process or accomplished via a Sequence job.

I've had good luck in the past with queries of this nature before there was an XMETA database by leveraging some of Chuck Smith's tools. Specifically the List all files and tables used by jobs in a dsx file modified to load the results into a table so I could query them easily. Yes, it's a bit old s...

anil411 wrote:If i change encoding from UTF-8 to ISO-8859-1 , The Data matches but we receive the file from external agencies.

I don't understand the "but" here. Are you saying you change the encoding in the job and it works? Or if you change the first element in the XML file?

Ah... good. Pretty standard ETL task, btw.

Seems to me you should be continuing to get IBM's help with this.

Create multiple csv files, limited to 1M records max each.

That sounds like a support question to me. Of course, someone here may have direct experience with it but in your shoes (while you wait) I'd ping support.

Agreed. I had assumed any research would show that it was bundled in nowadays after being patched into the older release. Guess not.