DSXchange

Each DataStage developer should have dstage (or whatever you've called the group) as its primary group, and this should also be the primary group of dsadm. For project-based security, there is a separate group associated with each project. Eponymous group names are good; a project called CDR could h...

Famous last words...

Most databases use an NR1W locking scheme (N readers or 1 writer). Therefore, unless another process is updating the table, you ought to have no problems whatsoever. Some databases, such as Red Brick Warehouse, implement "query priority concurrency", so that an update can be suspended while a snapsh...

A 32-bit integer (int32) can not fit into a Decimal(8,0) column. These integers can have up to 10 digits - the largest integer is 2147483647. That's why DataStage issues you with a warning. What you're trying to do may have catastrophic consequences for large enough integers (such as a segmentation ...

There is no Analysis Server running at IP address 10.119.2.13 is a very clear message.

You have either not installed or not successfully started the ProfileStage analysis server on the machine whose IP address is 10.119.2.13.

Welcome aboard! :D I'm guessing you chose to use an external database (SQL Server?). An error creating an index usually suggests permissions or privilege problems. An "internal" database is the one that AuditStage uses if you don't specify an external database. I'm not 100% certain, but I think it's...

I didn't tell you this, but a complete product revamp is coming sooner than you think.
I will take the Sergeant Schultz defence to any further questions on this.

The original design is perfectly valid and will give the same parallelism as the newer design. In the original design, because there are no interdependencies, the separate Transformer stages - and therefore data streams - will be handled by separate processes. On a multiple CPU machine, it's highly ...

Mostly we rely on the SQL*Loader bad file.

XML PACK is not free.

Given (reading between the lines) that there is no stage type to meet your requirements, the next best thing to do is to craft your own control script, and to use a Sequential File stage to generate the data file, and FTP (either an FTP stage or a shell script) to transfer the data file to the Windo...

Welcome aboard! :D Are all these tables in the same database? If so, and no transformation or checking is required, bypass DataStage and do the whole thing within Oracle. INSERT INTO table4(column_list) SELECT columns FROM table1, table2, table3 WHERE join_criteria ; Otherwise, use the same ...

The easiest way is to create a routine, and use the result of the routine (loaded into a stage variable, so you only need to execute it once) to constraint the output links. Here's one I baked earlier. FUNCTION IsDateStr(Arg1) * History (most recent first) * Date Programmer Version D...

You need to recompile the job to repopulate its RT_CONFIGnnn information, including the JOBPARAMINSTS record.

Ultimately, whether you do the joins in the database or (to DB2 tables) in DataStage, the same amount of join work has to take place. The same amount of work still has to take place if you use hashed files, but these can be cached in memory (if not too large), which is where the throughput gain come...