DSXchange

Use a Server job with the Folder stage as the target.

Bad Idea. What exactly did you remove - a file that ended with a .ds extension? If so, that's not the dataset but rather the dataset descriptor file that has all of the metadata, including the location of the actual dataset file(s) which could be located across multiple nodes. Deleting things manual...

You... extract the data from Oracle, convert it to XML using the XML Output stage and then invoke whatever loader utility Siebel uses. :? Surely, however, you've been given more information than that? What the XML needs to look like? How to load that into Siebel? Kind of retarded if the answer is no...

Your linked post shows your answer as well. Not sure what else anyone could add.

As long as there is an H: drive there, I would assume you'd be fine.

Yes, expected. Declare your output field as Char and then use DecimalToString() to get whatever output format you desire.

If you checked the ulimit values manually from the command line, that won't help. You need to check from a job to get accurate information - add the command as before job ExecSH and post the results.

You've got another post for the backup question (as you should) so won't address that here. And when you say "still is down" does that mean that you still cannot start it? If so, you'd need to go through those logs again and see what the issue is now.

Do you have the same issue if you do the import from the GUI, i.e. from the Manager?

Yes, we're glad, too. And figured it would come down to a namespace issue, they can be quite a PITA at times.

None of the repository tables, be it the new XMETA relational database or the 'traditional' Universe repository are documented. What I'm assuming you are referring to is the Universe table DS_AUDIT which contains the type of information you are looking for. A search for that table name here should t...

What O/S? It sounds like you've hit a kernel parameter limit with regards to the number of open files.

Next time that happens, Reset the aborted job and let us know if a new log entry is added labeled 'From previous run...'. Sometimes the details that it logs are the only clue we get as to the exact nature of the failure.

And you're certain of this why? Have you made the change it suggested and still have the same error? Also, have you ensured your kernel parameters meet the minimal values shown in the documentation?

Not really.