DSXchange

If you set your commit frequency to 1 you will avoid deadlocks, won't you?

VC - I realize that bug; but what I am saying is use the row generator to generate just one row with a dummy column which you can then throw away, but output that through a transform stage where you do generate your correct date column.

The parent of that process is pid 815260. In the case of a "defunct" process, it means that it has sent a message to it's parent that it is dying, but the parent has not yet acknowledged that (see, even computers can be in denial). This can happen if the parent is in an uninterruptable state such as...

With @INROWNUM and @PARTNUM you can concoct your own unique identifier.

Yes, DataStage can do this, but not with a simple click. First, you need to create a list of elements, in your case the months, i.e. 2007/6 and 2008/3 would result in a string "2007/06,2007/07,2007/08,2007/09,2007/10,2007/11,2007/12,2008/01,2008/02,2008/03". Once you have this string you can use a p...

I am not sure why you didn't use a row-generator stage to feed a transform stage which would produce your date values in the PX job. Server jobs don't have the foggiest what a "date" type is -> the parameter is treated as a string and you need to do an appropriate ICONV() on the string to get the in...

In the Float metadata have you enable "allow all zeroes"?

Is your userid "user1" or are you in group "group1"?

I'm glad it worked, perhaps you could also mark this thread as "Resolved" in order to close it off.

IBMDS - how many digits of precision? The message,

Type mismatch for field 'RSK_NO'. Fatal error

is pretty clear that somewhere there is a discrepancy. How did you determine the Teradata DDL?

No, I cannot.

The close command is just an SQL command, so you need to check your database to see what options you have there in order to "shell out".

No, you cannot do that in PX. What about doing it through the close command in your SQL? I haven't tried that before, but it ought to be programmable.

What (exactly) have you declared the column's data type to be in Teradata? And what data type in DataStage?

I would start by looking at the Oracle installations and seeing what the configuration differences are and seeing if any of them might be pertinant.

what is a "n/w issue"? If you can see the appropriate lock in the cleanup resources window why not get rid of it there?