DSXchange

On a fixed width file the Display setting sets the maximum width of the data that can go into that column and will pad accordingly. I hope you're using a delimited file, in that case you do not have this limitation. The Display setting only pertains to viewing data, as far as writing data to the fil...

That whizzing sound is something again going over my head. At which release of DataStage are you allowed to create PX jobs on Windoze? I haven't seen this one yet, is it 7.5? Because, as far as I know, only connecting to a DS project on a Unix server with a PX license allows you to create or even im...

I think you could modify the routine in m post and change a parameter value in the current job. This won't work if the job parameter is part of a filename AND the stream is continous (without a passive stage breaking the flow) So, if your job looks like this: SEQ --> XFM --> SEQ, then the target se...

Make sure you use the "Get SQL Info" button on the ODBC stage so that it properly registers the information about the ODBC driver in use. You'll see the Quote character text box right next to it probably change values now that it knows more about the ODBC driver.

I'm still trying to figure out how folks are designing Parallel jobs on the Windows platform. Even if you can't compile them, how the heck are you even designing them????

Search the forum, we've covered using the DS BASIC APIs in a user-written after-job routine to dump out log messages generated in the run to a text file. In fact, this question gets asked almost every day.

Hey, don't bust on Nag for splitting his hash files. If you're attempting to stay under a 32-bit limitation on either the filesystem or the hash file because of volume, splitting your hash files up into separate files can yield a lot of performance boost. Get ready to learn some new tricks here: 1. ...

Every DS job design has an number assigned to it. Every stage and link has a number assigned to it. To get the variable (array) that contains the entire row contents, you have to reverse engineer the job design information. To know the array name at runtime that contains the primary stream row in a ...

Please search this forum, we just last week had a long discussion over Y2K issues with Julian dates and the DB2 plugin. Your solution is discussed in that thread.

Yes but in different ways. If your data is two tables in the same source instance you may wish to just use SQL to solve the problem. If your data is in two sequential files you can use the Merge stage. You could also build hash files to accomplish what you wish, using a UV/ODBC stage with multi-row ...

You CANNOT use the WITH trick for ordering data if it's an externally pathed hash file. It has to be a local account, so remove the fully qualified directory path and use a blank project/account name. It will then use the VOC pointer entry to find the path to the hash file.

Maybe you should post your actual data instead of the emp table example. That way we can talk about apples instead of looking at oranges.

It does but that's not your problem. Your problem is that your design is wrong. You've assumed that the SQL is joining one way, but you need to cogitate about relational theory and realize what you think is the stream is really being handled as the "reference" vice versa. Change your stream link to ...

cat file_a file_b | sort -u > file_c This is an exact match bitwise, " x" and "x" will not be the same. It gives you unique rows from both files, a full outer join. Using a job, you can funnel/link collect all rows together and aggregate and group by and use min or max to collapse rows together. Th...

Depends on what your header information contains. If it is supposed to contain information about the row count, you have to create the header row after you created the file. Why not let DataStage create your file, then use an after-job script call to create the header and trailer record and then ass...