DSXchange

Hashed files can be accessed as ODBC sources - performance is going to be quite a bit slower than accessing them in server jobs.

Somehow the descriptions in this thread make it seem very confusing to me. 1. You have an Oracle Field "X" which is nullable 2. You read this field into DataStage, where it is still called "X" and is marked as nullable 3. You perform a NullToEmpty(X) and receive an error. Is this...

Did you define a system level ODBC named "FINDEV" in Windows? If yes, try using MSQUERY or similar to see if they can access "FINDEV". If they cannot connect then the problem is not within DataStage.

NewColumn = in.Field[1,1]:FMT(In.Field[2,99],"3'0'R")

I should also have added that doing a "dump" of the dataset to a sequential file would only allow SQL-like queries when using an ODBC stage on that file. Rather complicated when DataStage itself allows some extended processing on data...

I'll side with Craig on the NULL in sequential files question. Try writing a null to a sequential file in DataStage and then reading the null.

I don't have datastage accessible now, but you can open your DataStage Developers Guide and look at the modify stage chapter to get the list and syntax of the functions.

Are you generating the CREATE TABLE? I can't check right now, but seem to remember that you can view the generate DDL in the stage and then you could either post it here or manually try to run it in your favorite query tool.

If the database has indices on the join criteria and if it is less loaded and perhaps more powerful than the DataStage machine then it would make sense, from a performance point of view, to perform the join there. Usually this is hard to guess and the only way is to perform an empirical test to comp...

The documentation (which I don't have at hand right now) for the MODIFY stage will show the explicit conversion functions that you can use to avoid this warning message. Another option is to use message handling to demote or suppress the message - but I would rather avoid doing that and recommend us...

Are all 8 tables in the same database? Is the database server heavily loaded and is it a powerful box?

tehavele wrote:...main_program: Driver doesn't support the datatype. DataType Code = -9
main_program: Could not get mapping for one or more datatypes. ...

What datatypes are you using in your table. One or more of them are not supported.

Could you cut-and-paste all the error messages - there should be more information available than what you posted initially.

Yes, that makes a significant difference, as you have noted.

If I understand your issue, you want to ensure that if records going to A are rejected during the write phase that they don't go into B? In that case, you can put a reject link out of A and then make that the reference stream to a lookup stage coming out of the transform stage and heading for B. All...