DSXchange

Think of it this way. In a text file there is only one data type - text. The SQL data types you add to the columns grid are merely documentation when in a Sequential File stage - they indicate the most likely SQL data type should you ever change the Sequential File stage to a stage type that uses SQ...

The real time interface (RTI) allows you to set up DataStage jobs that can run as a service in real time. For example, a data entry operator enters name and address details then, when complete, an RTI service is invoked to validate the entered data. This might even go so far as to use Quality Stage ...

What stage type are you using? Which database?
Are the columns participating in the joins indexed? (If so, then DISTINCT should work very well.)

PS If this really is a question about PX (parallel jobs) can you please post it on the Parallel Forum?

Do a run of your job from Director. In the Job Run Options dialog (where you enter parameters, etc.) choose the Tracing tab, select the active stage (Transformer stage) and enable collection of statistics. One extra event will be logged; it will tell you what proportion of the time was spend doing w...

You can't create a DSN on the fly unless you know what the DSN has to refer to, the driver it has to use, and so on. If you know all this, you can create DataStage BASIC code (probably a job control routine, or a Routine to be called from a Routine Activity in a sequence) to edit the relevant files ...

Welcome aboard! :D Does your nickname mean you're an Aussie working in the Netherlands? There's no DS390 equivalent of the XML writer stage. So, yes, the only way to do this solely using DataStage would be to generate, for example, a delimited flat file on the mainframe and use an FTP stage to send ...

If the other application writes to stdout, you could specify that application as the "filter" in a Sequential File stage that reads the data - the downside being that the application would be started by DataStage. But it's really going to depend on where the data are in the application, and how data...

All responders (except me) have assumed what the OP meant by "locked". Beware of assumptions. As you may already be aware, or soon enough will become aware, they are highly dangerous things in an ETL environment! Let me also make the point that killing a process will rarely release some kinds of loc...

That's a real pain! How do you even get to know the names of the databases to which you will need to connect? You need this before you can even create data source names. Could be an interesting process to try to automate! DSNs can usually be created via a simple edit of the $ODBCINI file (some diffe...

The answer to the original question is on page 2-12 of the DataStage Enterprise Edition Parallel Job Developer's Guide . Performance implications can be inferred. Preserving anything involves "no change" and therefore no processing (though note that the following stage may ignore the request to pres...

There are separate chapters in the Parallel Job Developer's Guide for each of these stage types. Make a grid of the things that you aren't sure about, then find out how the stage types differ in these respects. Do this then post again if your question has not been answered by doing so. If it has ans...

You know that you can do it by choosing Options in the Export dialog and choosing "Individual Component" as the selection method. The actual export is driven from within the Manager client (or, more precisely, from dsexport.exe which is called from it). This interacts with the DataStage server throu...

Is there a finite number of these databases, or is it completely arbitrary? In the former case, you could make the DSN a job parameter and control it from a sequence or from job control code quite easily. What you do about differing table metadata is quite a different question and one which is not e...

NEO wrote:Can we do a lookup with an ODBC stage??

Yes

NEO wrote:what could be the difference of doing it with hash stage and a ODBC stage.

ODBC generates SQL which has to be processed by an ODBC driver then by a database server to access disk-based tables; hashed files can be loaded into memory. Hashed file stage can only be used when there is an "=" lookup on the key; ODBC (and other SQL-based stage types) can implement other kinds of lookup, such as a date within a specified range.

NEO wrote:Do I get any performance issues if I do a lookup with ODBC stage??

Yes, huge differences, particularly if hashed file is cached in memory.

1. Write them to a hashed file with write caching disabled. All updates to a hashed file are destructive overwrites. Then load them from the hashed file into your target database. 2. Use an Aggregator stage, specifying the set function called Last on the column(s) in question and grouping by all oth...