DSXchange

:!: Split from this older topic so you can control your own destiny. First off, confirm for us you are still talking about a Server job on a Windows server as those "SPIN" variables depend entirely on your operating system, from what I recall. Also let us know what DataStage version you ar...

lmoose1 wrote:We just installed 8.7 on Windows Server 2008 R2 Enterprise SP1 and are encountering transformer compile errors, too. We are using Microsoft Visual C++ 2010 Express Edition for the compiler.

First question - is that exact version of the compiler supported for your release?

Auto-purge will only happen when the job finishes. Surely you could take it down for a minute or two in the middle of the night each night to allow this to happen? Otherwise you'll need to write something that would run at a certain time and delete entries up to a certain point. I personally wouldn'...

You could use a Hashed File for this, keyed on YYMM & EmpNo. When the lookup fails, write out a new record with the key and the hours in the proper days, otherwise sum the hours and rewrite the record. When complete, source the data back from the Hashed File. You could also use an Aggregator, gr...

Let's start by getting you into the correct forum... hold on a sec... there.

Job logs won't auto-purge on an abort, unless that's a change...

What makes you think that? Recovery would mean from a backup or (hint) from a copy you made before you changed the original. You could post the contents here, that way someone could take a look and see if anything looked... damaged.

That would be a 'bulk load'... and as far as I know, that would require you to build a load-ready file and transfer it to the SQL Server box, then trigger whatever passes for a bulk load in SQL Server.

Can you not make the job 'always on'?

(and off we go to a more appropriate forum)

Sort in the proper order and then keep first or last, depending on said order. Hash partition if running on multiple nodes. Of course, you could also use an Aggregator but that wasn't the question.

Several things need to be said here before your question gets answered. 1. Welcome. 2. When you have a question, start a new post. Please don't just reply to whatever post you happen to be reading at the time. 3. I split yours out to the above noted post of your own. 4. Because of that, I had to gue...

However, the issue is having a stored proc stage in the job would mean that every time the job is invoked it will login to the database regardless of data being pass through the link or not. which is what we are trying to avoid here. You are trying to avoid the unavoidable. Every database stage in ...

So Roy, is this Resolved or Worked Around or are you still futzing with this?

I don't see how "disabling run time column propagation" would have anything to do with the fix but glad you got the rest sorted out. As Ray noted, just saying that all three jobs had identical parameters in them didn't help, they actually had to be identical.

You're just moving the decimal point, so divide by 100 after conversion.