DSXchange

Where are the locks coming from? If it is from the job itself, have you tried running on a single node as previously noted? That would eliminate the target as the source of the locks but would still lock if they are being generated on the source side.

... column names you would have to manually hard-code and add to the flow before pivoting.

What happens if you change it to two anonymous blocks one after the other rather than embedding them like that? Or run both scripts inside one anonymous block?

suryadev wrote:please suggest how to retrieve the whole tags?

eostic wrote:When it ends in a slash, you get the whole tag.

So your desired output would be... Empl_rec, Daily_Change_Ver_Num, Changed Data 03456, 3, "Empl_Status, Leave_of_Absone_Date, New_Salary" 45633, 2, "New_Address, Empl_Status" Correct? And your changed data field, is it in fact one field containing a comma delimited list of the ch...

You'd have to clarify what "Col_3 is different across the records in the group" means. Can you supply examples with real data, input and output?

You don't really need to Sort or leverage a Key Change column, just run the data through an Aggregator. Group by Col_1, max(Col_2) and do whatever is appropriate for Col_3... depending on what "Data" means.

Welcome.

Do an exact search here for "Controller problem: Unhandled failure" and you find 50 other discussions on this subject.

Well... classic first question would be, if you run this on a single node does the problem go away? Secondly, have you had your DBA monitor the session when it is hung?

Can we get a clue as to what the after SQL actually is? I'd also be curious what makes you think it is being skipped.

Just an FYI, you asked this in three posts - one is just fine. I removed the other two posts.

FYI - You are able to go back and edit your original post to correct the Job Type...

And so off we go!

Can you specify what kind of locks you are seeing? Posting the actual error messages should help, I would think.

I'm not sure I'd be relying on the logs for information like this. Is this something you could specifically store outside the logs so you have more control of how it is stored and how long it is retained?

Is this a parallel job? I only asked because you posted in the Server forum.