DSXchange

32 in parallel may saturate the machine, and therefore be self-defeating. Opt for a compromise of N in parallel, where the machine has N+1 CPUs.

Add a message to your code that is logged after your code exits from the loop. You will then be able to confirm that the loop has finished. Note that OPENSEQ takes the ELSE clause if the file unit is opened successfully but the file does not yet exist. You need to test the STATUS() function for this...

Any time you get a "library can not be loaded" error the first thing to suspect is the environment variable that defines the library search path. Depending on which UNIX you're on this is called LD_LIBRARY_PATH, SHLIB_PATH, LIBPATH, or something similar. Make sure that it is set up properly in the $...

Is qsrtmngr running? What happens if you try to import a table definition via the QualityStage plug-in? In particular, what error message appears?

Put the SELECT statement in one stage, run the results through a Transformer stage (even if you do nothing to the data) and have the final stage perform the INSERT (or bulk load). Create another job that updates the control job; similar design: SELECT in one stage, Transformer, update in final stage...

Here's one (it only takes one pass, but is executed by an external job running the loop - you could wrap the loop around this code if you preferred). FUNCTION FilesInDirectory(Directory, WildCard, OutputFile) * History (most recent first) * Date Programmer Version Details of Modifica...

It's a connectivity error. The "handle" (a variable used in the OCI to manage the connection or the statement) is invalid. Possible causes include improperly specifying the connection string or database name, or a connection that has been lost since first being connected - that is, a network failure.

Is there a mechanism in the database table that allows you to answer the question "5 days worth"?
Then surely the simplest mechanism is to use an appropriate date difference function in the extraction SQL. For example

Same solution as in previous thread with the following enhancements.
(1) Create one file of all the input files, using the TYPE command. Process the combined file.
(2) Change the constraint expression to Mod(@INROWNUM,6) = 0.

But Ray, doing with simple unix commands will simplify the job to a great extent. Using the stream editors like sed will simplify the preparation of the sequential file without having to go for a job. If you are calling DSJobReport from the After-Job subroutine, you can very well use DataStage Basi...

Parallel Extender is a part of DataStage Enterprise Edition. :D

For everyone else: Input link has one column, called DSLink3.Line. Delimiter character is set to 000. There are seven stage variables, all initialized to "" TrimLine sv1 sv2 sv3 sv4 sv5 sv6 Derivations for the seven stage variables are TrimLine Trim(DSLink3.Line) sv1 If @INROWNUM = 1 Then sv...

Click on my identifier; email is enabled from here. Please do the same in your profile; I will send you the image of the working version. Which is exactly as I described earlier.

Welcome aboard! :D There are several solutions to this problem. If you've already loaded the master records, then you can load a hashed file with each master key. A non-key column could store the "next value" for each master key. Indeed, you could even load them into the SDKSequences hashed file use...

You can send the DSX to Craig. He obviously has some spare time!