DSXchange

14347 could well be a DataStage internal format (12 Apr 2007). Try applying an appropriate Oconv() function to get it into the format required in the SQL.

Bad code! You go on to attempt ClearFile even if the file is not opened! * Arg1 contains pathname of hashed file. OpenPath Arg1 To hFile On Error ErrorCode = Status() Call DSLogWarn('Error opening "' : Arg1 : '".', "Before Job") End Then ClearFile Arg1 ErrorCode =...

Were the working ones compiled in an earlier version of DataStage and have not been recompiled since?

Using a Hashed File stage it is only possible to do exact match lookups.

A "closest to" is a complex query, involving a partial key match and a MIN on a date difference, but could probably be performed using a UV stage. Just don't complain about the slowness.

To verify parameters look in the "job starting" event in the job log to see the parameter values that were passed. If these are ALL ok, revisit the encryped parameters. However, this is not your problem. You got an error code and message back from Oracle. Therefore you successfully connected to Orac...

Craig's questions are pertinent; without knowing what stage types you are using and data volumes you are dealing with, for example, it would be imprudent to provide advice. What is the basis for the merge? Are the source and target tables in the same database instance?

When you searched the forum for this error message text (perhaps ignoring the specific job number), what did you find?

Why?

A Pivot stage will do exactly what you require.

With a hashed file, the technique for a horizontal pivot is to load the non-pivoting columns in a value mark delimited dynamic array and to UNNEST that when extracting.

Look in the Built-In\Utilities category.

There was no need to shout. The error message is "failure to connect". Unless you can connect successfully you will not be able to import metadata or do anything else. There is something wrong with your connection request - it might be user ID/password, it might be server name, it might be a misconf...

There is no command or file in the QS environment to see what command was executed. However the DS job log will report what command was sent - which might prove to be of some use.

You must have the Visual Studio .NET C++ compiler.

No. Your requirement (original post) was SEQUENTIALLY.

File Sets and Data Sets are read in parallel.