DSXchange

Best practice is never to abort, to remain always in control.

Do there really need to be five columns in the lookup?

If you are simply determining existence, then a key-only lookup would suffice. If you are using only one returned column, then just the key and that column are all you need.

Welcome aboard. :D

Is the run time library ARDTRTLB installed on the mainframe machine? Are you using any custom routines?

Don't put it in there in the first place, and you won't need to remove it. Call DSExecute("UV", "COUNT hashedfile", Output, SystemReturnCode) or HashedFileName = "hashedfile" Call DSExecute("UV", "COUNT " : HashedFileName, Output, SystemR...

If you can run dssh in the project directory on UNIX you have already established a telnet session.

DataStage does not provide a telnet service on UNIX. The only reason it does so on Windows is that there is no native Windows telnet service.

Start by designing exactly what it is you want to do; what the requirements are for input, precisely the algorithm(s) to be applied, the metadata design for the output(s). Then research the very rich set of stage types in the product, to be certain that you can't do it with some combination of those...

It's exactly as the error message states. In your source data the value 2,222,222,222,222,220,000,000 occurred. You tried to load this into a column whose data type is declared as Decimal(10,0). It quite simply doesn't fit. Choose a more appropriate data type, for example Decimal(40,0) and the error...

Kumar has two problems. You can't execute an unqualified uv command from the bin directory on UNIX unless . (the current directory) is included in your command directory search list (that is, the PATH environment variable). You can't successfully execute a uv (or dssh ) command in a directory that i...

That's not what DataStage does. I, like Craig, would make this point very, very strongly, and push back hard. However, if you have a true PHB, you could create Transforms or Routines that applied HTML tags. For example a Transform expression might be "<b>" : %Height% : "</b>" Dat...

Any errors or warnings? Can you successfully view data in the second ODBC stage's input link? (An informational message reporting "no data found" counts as success.) Are the row counts on the second stream reported as 0, or does this stream not execute at all? Can you create a second job only contai...

The NLS map for DataStage Hashed Files must always be NONE. This is because DataStage Hashed Files operate entirely within the DataStage environment; any characters needed are mapped on the boundaries between the external world and DataStage; remapping can only cause errors. Of course, if you're acc...

You missed the one combination that works, with the semi-colon inside the quoted string. uv "select * from DS_JOBS where NAME = 'PSEXPSLSE';" Ascential/IBM provides no documentation as to the other tables in, and structure of, the Repository. This is a deliberate choice; they reserve the r...

DSU.getBatchRunDateVal is a BASIC routine; I guess you're invoking this from a BASIC transformer? The error message shows that you have compiled in trace mode (and are using tracing?). It appears that an attempt to clean up a temporary file (RT_SC4771/jpfile.6A6F625F4A4F42676574446174655F4461696C795...

I don't believe there is any way within DataStage.

If the records contain timestamps for created and updated, you may be able to determine these counts with SQL queries.

DataStage looks for the uvodbc.config file in the project directory. Only if one is not found there does it look in the $DSHOME directory. And if it's not there either it checks in /etc before giving up.

It's not an issue; that's how DataStage works.