DSXchange

That should be good enough. How many times did you stop the job when it hung? There might be live threads on your target table. Ask you dba to get rid of all such threads, run your job again, after fixing the data, and ask your DBA to monitor whats going on. Also clear the status file before you try...

What is the transaction size that you have specified? Also about the missing quotes, you have data with missing leading quote. Fix the data.

I_Server_Whale wrote: There you go, DSguru2B just spoke my mind.

O yea. I have Ray's RMM stage with me.

Your not wasting anyones time. You will always have someone or the other to reply to posts. As Whale put it, playing is equivalent to learning. Didnt mean to discourage you with my comment up there. A friendly piece of advice, you have manuals, go through them. They will give you a good understandin...

Well, write your max value a sequential file. Build a sequence job which has two an command activity stage. Pass the command type your/fully/qualified/filename Have a link coming out of command activity stage to a job activity where your calling your job that generates the keys. Open the job activit...

Tell me about it. I havent used row splitter and row merger in ages and i brushed up my knowledge on them

vijaykumar, are you just playing with different stages

There have been a few folks that encountered this problem. Search on the error message.

I suggest starting a new post with a link referring to this particular post for referrence. This way we will know job type, os type and other information that will help others provide a solution.

Cool

And Craig is right too. I just added ABS() because our ghost friend mentioned larger date from smaller

What if I want to call a userdefined routine. Can i use the same format with DSU. instead of DSX. ?

Ghostraider, (looks like you a big fan of ghostraider, the comic hero), Ray has a point. Remove ABS() and you will get negative difference if second date is larger than the first one.

Sure. Use a transformer to split your columns. Pass the gouping keys and sum to odbc and the remaining to the sequential file stage.

Thats like a more general processing question. Search for it on google. Let us know what you found