DSXchange

Can you be more specific as to what "the value fetched is not a unique value" means, please?

Also, SQL Server does in fact have SEQUENCE objects unless your version is older. You could possibly also use an IDENTITY column and save all this.

Read what the DaysInMonth() function actually returns, it's not what is needed here. While you probably could cobble together a solution using it, I was suggesting a function like MonthDayFromDate() instead.

Yes... should be the quickest way to a solution. Note I didn't say "quick", but rather "quickest".

Sorry but that isn't quite what I suggested. You need to offset it by the current day (plus 1 and made negative) not the current date.

Of course, because you simply added 1 day to the current date. Re-read what I wrote... after I fix a very important omission about flipping the sign.

You might want to check out the DateOffsetByDays function. If you take the system date and "offset" it by the ((current day + 1) * -1) days, you'll get the last day of the previous month.

Seems to me it is time to involve support, it you haven't done so already...

Never used that function, so don't know if there are any restrictions to using it but your examples don't seem to match.

Where you say "To be specific" there is a "*100" on the RHS that isn't anywhere else.

Based on the wording here and the linked post, it would seem they are already using an encrypted password.

ksfun,

Split you out to your own post, and since it's been 11 years since you (or at least this userid) has posted here, I had to make a guess as to your particulars. If you could edit your post and update your version (etc) to whatever is current for you, that would be great.

I realize you said "unknowingly" but can you tell us where you were working and what you were working on / doing? Maybe that would help get us the information Ray was looking for...

So, out of curiosity, is your "timestamp" field actually a string?

So what actually is your "workaround"?

See, I told you it would help you figure it out.

As Arnd mentioned in the linked thread, the use of the CAST function in your source SQL creates (as far as DataStage is concerned) a new column that is nullable. It doesn't mean you have null values in the data, just that the job thinks there could be, which from what I recall triggers the error whe...