DSXchange

Import as in the operator, Roy... i.e. read the file.

Are you running the job on multiple (as in two in this case) nodes?

You'd need to explain exactly what "fails" means for help with the second question.

"The database is damaged"? Not really sure how a job would cause that error. Have you involved your DBAs, done any troubleshooting as noted here?

Non-specific answer - persist it. That or run SQL against the table to get the current max value + 1 and pass it into the job as a parameterized starting value.

ps. You really should be using meaningful names for your stage variables.

Marking the topic as resolved for you. You can always mark any of yours yourself and also re-open them if you need to.

I'd suggest you contact support, see if this is a known (i.e. patched) issue.

So... what problem still remains?

:? Honestly, I'm not following and I think it is going to take "eyes on" to understand your issue. Perhaps if you took screenshots of the two job variations and posted them here? Might give us a better starting point for this conversation. We might also need to see inside the transformer b...

Can't? That's a pretty severe limitation if you can't build work tables. IMHO.

You need quotes around your string and there's no square brackets needed.

Also there's no need to quote everything all the time, rather than 'Reply with quote' simply use the Reply to topic link. Save the quoting for when it makes sense.

StringToTimestamp. And you'll need a time as well, even if it's all zeroes.

Interesting... not sure I would have thought of that approach.

By... running the job on a single node. Hence the suggestion.

Do you know what that means and how to accomplish it? Have multiple APT_CONFIG files, one of which one has a single node defined it in?

One of the dangers of running A/V and other programs of that ilk on an application server. You should be able to tell it to exclude the directories that DataStage lives in and then still be able to run it.

So you want to output only the missing rows?