DSXchange

When writing to SQL Server, on version 7.5.1A, ODBC stage cuts off the first character for character columns and DRS stage cuts off the last character. I find out DRS stage better. It is more powerful.

Yuan_Edward, I cannot use Floor my integers bigger than Datastage will allow.

The error being returned is:

"Numeric string expected for returned value from function call
'strip_whitespace'. Use default value"

TempVal: StripWhiteSpace(Link1.IncomingCol) * 100 Data type of IncomingCol is char. If I do: Left(TempVal, index(TempVal,'.',1)-1) it gives me the error: Numeric string expected for returned value from function call 'strip_whitespace'. Use default value. However, if I do: Left(Left(TempVal, index(Te...

Ok. My apologies. Turns out that this was due to that column's length. I needed to Trim and reduce the size by 1. Thanks ArndW for your help. Really appreciate it.

I have a 2 node configuration. The partitioning on the transformer is set to Auto.

Here is an interesting observation. When I send the data directly from the join stage to a sequential file, it is fine. However, when I send it through a transformer stage, the data has duplicates. This is despite the fact that, there is absolutely no transformation being done in the transformer sta...

Result is the same with 1 node config file.

Running it with a single node configuration file is not an option as I can't do it in production, so I didn't try. I did see one post here about converting running all previous stages sequentially which I did but to no avail.

I have 3 files going into a join stage. I am doing has partioning and sorting on Col1 on all 3 input links in the join stage. File1: 200 rows (Col1) File2: 50 rows (Col1) File3: 50 rows (Col1, Col2) My key column is Col1. Col2 has sequential increasing distinct values. After the inner join, I get 50...

I take that back. I haven't found the solution. After I converted everything to integer in a transformer stage, the join still returns less value. I'll post some data.

I found the problem. It is due to join on a char field. The actual numeric data inside is smaller than the field width in all cases and this is causing the problem. Is there any way to get around this? Should I pad blanks on the left?

I have 2 links joining 10 rows each. I know that they are identical on the key column so that is definitely NOT the reason. The key column is char but has numbers like 1234. The number of rows I get after the inner join is less than 10. Could the positioning of the numbers in the column be a reason?...

But how would you tell QS this:

"Interpret A as Ave in this situation"?

/ is not in my striplist but I can add it with an escape. However, 34 1/2 is the street address and that's how I want QS to see it in my override. So if I had the / in my strip list, would my override be: HN1PD1SN1SN1ST1 I guess I am going to run this through, find the address and see how QS treats ...