DSXchange

DataStage includes optional XML reader and writer stages. You may need to install them using the DataStage Package Installer. The XML reader stage sounds exactly what you require.

Ray Wurlod
Education and Consulting Services
ABN 57 092 448 518

Earlier this year in Mumbai I implemented a similar function, though it retrieved the parameter values from an Oracle table rather than from a text file, and they had to take into account separate instances of jobs that ran in multi-instance mode. I don't intend to post that code here, since it was ...

It's almost 100% certain that you have NOT gotten a successful installation. Was this a clean install, or an upgrade from an earlier release of DataStage? Is the UniVerse RDBMS also running on the machine? You may even be able to connect to a partially installed DataStage server. Were you asked abou...

L4 is Transformation Rules Compliance. Here you simulate transformation from source data to target data, and perhaps search for alternative transformation logic that might improve the end results (for example whether particular cases (NULL?) have been handled "properly"). You also verify, of course,...

This is exactly the kind of job that INTEGRITY does. It specifically uses probabilistic matching algorithms. I'd advocate using the correct tool for the job (INTEGRITY) rather than trying to do the cleansing job solely with DataStage. You can use DataStage to prepare data for the INTEGRITY product -...

This is quite an old version of DataStage (4.2 or earlier). Nowadays the installation script offers to shut down the services for you. If you're on a Windows platform, open Control Panel then find the UniVerse applet. This opens the UniVerse Control Panel from which you can click "Stop All Services"...

When you are running multiple instances of server jobs (which, from your context, is what I assume you are doing), you are actually generating multiple instances of the job design, including any contents of shared containers. Thus, at run time, they are running completely independently of each other...

There are a few ways to achieve this. I am assuming you're asking about server jobs - techniques are different in parallel and mainframe jobs. The easiest depends on your input being sorted, then you can use stage variables to detect the duplicates (no change from previous row) and the results of th...

The usual reason for using a stored procedure is that you already have a stored procedure that does exactly what you want, and don't want to re-invent the wheel. In this case it is eminently sensible to want to use the stored procedure. To achieve this, you must first have imported the stored proced...

When you try to connect to a DataStage project there is a drop down list of available project names in the client's connection dialog. The message you're getting indicates that the project name you're specifying does not exist for some reason. One possible reason is that you've mis-typed or mis-case...

This is the kind of task that the MetaRecon product performs. It profiles data not by looking at the metadata but by looking at the actual data and inferring what the metadata should be. Find out more from the Ascential web site, or contact someone in Ascential, or visit the sister site of this one ...

Restartability always carries with it a burden of needing to stage data on disk. You have to design for this, since DataStage is intended to keep data in memory as much as possible (for speed). Best practice has evolved through people's experience; I think most would agree that a hierarchy of contro...

Import the table definition from Oracle and see what data type the column is reported as being. This will probably help you to determine the most appropriate data type to use within DataStage. There is a list of supported data types in the Server Job Developer's Guide (and possibly in the online hel...

Take a look at the SDK routines for data type conversion, for example DataTypePicComp through DataTypePicComp3. These will provide some clues. You may be able to avoid "re-inventing the wheel" by using these Routines directly!

Ray Wurlod
Education and Consulting Services
ABN 57 092 448 518

In a job control routine you use the GetParamInfo() function to retrieve the value or - or other information about - a job parameter, and you use the SetParam() function to set the value for a job parameter in an attached job. Full information on both of these functions is in on-line help. If you're...