DSXchange

My initial thoughts are that the slow speed is caused by the sort stages but these are required to ensure the earliest start dates are loaded first. Is it possible that it is fact caused at the DB level? That should be readily apparent... how much of the 3 hours are spent in the Sort stage? Once re...

Welcome aboard, but you really need to start your own thread on this subject, not jump on the end of this one. We don't know details like Server v. PX, operating system or version of DS that a new post would require. When you do start your thread, clarify what you mean by 'table'... database? Unive...

I tried using DSLogFatal in my transformer stage before but it did not work. I know we can use this subroutine in Afterjob subroutine or transform function but not sure we can use it in stage variables. Works just fine. All you need to do is call it and your job is dead, so not sure under what circ...

Or just set the transaction size to zero. That turns your load into an 'all or nothing' load and the records are only commited once at the end if the job completes without fatal errors. In other words, the job must abort for the rollback to occur. This means running the job with a very low Warning t...

Ok... first question would have to be... why? That being asked, there are routines in the sdk you can check out for examples - like the Key Management routines. You can also examine any job's log that creates hashed files, the syntax used to create them will be in the log. Keep in mind the fact that...

Actually, the error comes during compilation and specifically says: SP cannot currently be used as Reference Input So this either means it can't currently be used because the metadata is missing... or a more literal reading of the message would lead one to believe it is not 'currently possible' as i...

The concatenation wouldn't be an extra step per se. You would do it via a Before Job subroutine using ExecSH, so it would run as part of the job. However, the Merge stage can be used for this if set to Complete Set for the 'Type of Join'. On Kris's suggestion re: the Temporary Directory, you'll just...

Do you really need to 'merge' these two files? Or do you simply need to load them both, one after the other? If it's the latter, why not concatenate the two files together before job and then bulk load them?

Do you know what the maximum number of fields could be? If so, you should be able to setup the Sequential File stage to read the file using the '^' delimiter for all possible fields and telling it to set any 'missing' fields to null. At least you can in Server, assuming something similar is possible...

Thanks DSguru2B, that's a useful suggestion, but I cannot implement it because my target tables reside on a different database from staging tables. I only have access to create tables at staging level, while my lookup table is a target table. That really shouldn't be an issue. Can you not pull a li...

You need to properly configure DataStage for Oracle, something that is documented in both the Plug-in Configuration Guide and the OCI documentation. It involves editing the dsenv file to set $ORACLE_HOME and to include the proper libraries in your equivalent SHLIB_PATH. Once that is done you need to...

And do you understand why? Your null check was failing because the value was not a null - but rather an empty string. While they may look the same they are completely different - one is a known quantity (the empty string) while the null is an unknown value. The length check will return zero for both...

Post your code. And are you having issues running the routine or trying to test the routine in the Manager? Asking because you can't pass null values when testing in the Manager...

Easy? Not really, but it doesn't have to be all that complicated. You should check this post for ideas. Searches will turn up other conversations on the topic as you're not the first to have to tackle this.

Microsoft, .NET, Windows, Bill Gates... take your pick.