DSXchange

To me that falls into the "those that are inherent in the two platforms themselves" category.

No clue as I'm not privy to the inner workings of those functions. Perhaps Ray or Arnd might have a clue or you could ping your official support provider to see why that might be.

I don't remember the exact name of the variable, but the maximum precision is set to 15 in the uvconfig file and I would wager that's your issue. You can switch to "string math" for larger number or perhaps bump up the max but I'm unsure of all the ramifications of that.

There's no difference in the product itself other than those that are inherent in the two platforms themselves. The tool is the same regardless and behaves identically.

You can't use IN with a list of values with an AND between them, it may delete far more than you're thinking it will depending on the specific values you pass it. Think about it. If you want to delete 11/2008 and 12/2009 and say MONTH IN (11,12) AND YEAR IN (2008,2009) you will end up deleting: 11/2...

Another option, if your month and year are in the dataset, is to make sure your data is sorted and partitioned correctly and then issue a single row delete out a first link to a separate ODBC stage that executes and commits the delete for the pair before the second link sends any inserts. If you onl...

Yes.
Changed your IN to an = as that's the only way this would work with the way you've built your SQL - singleton values.

You'd have to collect that information in a separate step / job and pass it in as job parameters.

No, not literally "none", it should just be an issue of not running jobs that target the same table until you understand the issue. What kind of loads are you doing, exactly? Can you please post the actual, complete and unedited errors you are seeing? That way people who know may be able t...

Yes, I am using ''Automatically handle activities that fail option' as I want to fail the job for any failure [this sequence has lot of other steps/stages/jobs]. That's fine and exactly what it is meant to do. I should have been more specific and thought about this for a moment longer - your trigge...

Talk to your DBA about the "deadlock or time out" errors so you understand what they mean. Run the jobs individually in the meantime.

That's because you are using the 'Automatically handle activities that fail option' but your triggers should override that if you are using the right pair of triggers out of the stage: 'Conditional (OK)' and 'Otherwise'. Are you saying that you are and it still logs the warning? :? ps. A Wait For Fi...

I would use one of the the test operators, perhaps "test -e" or with -f or -s depending on how tightly you need to test it.

There's no direct equivalent in PX as far as I know. Something wrong with using the "Abort after x rows" functionality? Optionally, you could use a BASIC Transformer or Server Shared Container to call "UtilityAbortToLog", with all the caveats that would bring to the table.

You can't do this... it doesn't even make any sense as there is a possibility of the TRUNCATE being executed for every row. And as you've seen, once the stage "starts" and the table gets to be "in use" the truncate fails. Do the truncate in the OPEN command of the stage. And whil...