DSXchange

Yes, no, dev, test... figured the location mentioned was just a 'typo'.

A little lost with your answer, Ray. Why would the target connector have any clue about what the data looked like before/during the pivot? That seems to be the behaviour being reported. The transformer between the two should mask that and (it would seem to me) that all the connector should see is th...

Search here for the phrase "double convert", you can use two embedded Convert() functions to remove characters by providing a list to keep. For example, to keep all of the numerics: Convert(Convert('0123456789','',YourField),'',YourField) The inner convert removes all numbers from your str...

If the job needs to be updated, then yes.

Still haven't specified the error you get. And is the problem happening because you've told it to create the table rather than populating an already existing table?

You'd have to let us know the errors you are getting before anyone could help. And examples of your data couldn't hurt, either, especially to help clarify your "instead of values in transformer" statement which I don't really follow at the moment.

If you compile in dev and export the executable to production, there's no need to recompile in production as you already have what you need to run it there.

Right now, when the indicator next to someone's name says they are a Premium Member, all that means is they were at some point. It in no way reflects their current status. Which can be frustrating at times.

ray.wurlod wrote:This is a classic "fork join" job design. Search DSXchange for examples of how it's done.

You can certainly see enough of the post.

So... a report printed to a file. What have you tried? The short answer is you read it by defining it as one long string and then parsing out the individual fields you are looking for inside the job, propably based on the first value in the string. Lots of the informational will just be noise and be...

Obviously, you can't compile jobs with a Transformer without a compiler but you can run them. That would be the whole point of that option.

Lots of caveats, however.

LOL! And the correction is corrected... It does get confusing because the concept of which one is the "string" and which one is the "substring" reverses when you go into this IN/NOT IN territory. Which made Harish's posted example work out properly in spite of his posted code bei...

Try Resetting the aborted job from the Director and let us know that contents on any new log entry titled 'From previous run...', that typically has the gory details needed to troubleshoot those kind of failures.

To add to that, it is in no way 'mandatory' but more of a technique that is used in conjunction with schema files. And it's all in the manuals.

First, don't use "-wait -jobstatus" as they are redundant, use one or the other. Read about the difference in what status is returned to decide which is the one you want. Second, if this actually works but only fails occassionally, then I'd take a hard look at the parameters you are passin...