DSXchange

There are two many variables for anyone to know that answer, if you really want to compare / contrast the two methodologies build both. Me, I would assume the 'all in the database' would work out better.

As I tried to point out, the functionality it thinks it is seeing did not exist in your old version, so I'd be shocked as all get out if you can simply create a state file to solve this problem. Show us what you found that seems to be causing the issue - the name of the stage variable and the comple...

To me, this is a perfect example of when you should contact support and get some hand-holding, some personal help like I did. Not only to get the problem fixed but also to determine what caused it.

Does this mean that the left join based on two keys C1 and C2? Honestly, I have no idea never having worked with SAS. I've been making some assumptions in this thread, seems to me you'd be better off asking someone you work with who knows what exactly that translates to. The SAS code also implement...

You need to add new fields with the column header values in them to pivot along with their associated data fields.

Did you try searching for DBPARTITIONNUM here? There are a handful of other discussions that mention it, something there might help.

Post the exact trigger condition you are using and let us know what status the Director shows for the first job (not the sequence) after it runs.

ray.wurlod wrote:Four.

OK. That's what I thought... which means it doesn't really emulate the Informatica unconnected lookup but is akin to a 'connected' one, or rather four connected ones.

If you are trying to import the metadata for your csv file so you can read or write it, just use the Sequential File metadata importer, there's no need to use ODBC for this. Or am I missing something here? Are you trying to turn this flat file into an actual database table?

Not nearly enough information to help. What's in the sequence job's log, any warnings? What trigger expressions are you using between the Job Activity stages?

And I would suggest you always start the Sequence job and then, depending on the date check, decide if the rest of it can run or if you need to send your warning out.

Compile a C/C++ program.

As noted, multiple ways... do you need to worry about the year after a leap year?

Could be more than simply the placement. What exact version / fixpack are you running? From what I've read, that option was added after the fact.

Do you need to pass any job parameters to the command? If not, then yes you can leave that part blank. Of course, you could just try it and see what happens.