DSXchange

Vimali, can you use the "..." to get a drop-down list of directories & files? If you specify it this way does it work?

The values displayed by rows/second and number of rows processed are not accurate until the job has completed; the interim results are affected by several things and are seldom exactly correct; especially if the job runs in less than a couple of minutes. If you have buffering turned on for your hash...

Yes, both do the same thing. The DS.TOOLS command is a simple menu, and when you enter (2) it just calls up the command "DS.REINDEX ALL". The DS.REINDEX is a PAragraph, or script, that goes through all of the DataStage repository files that use indices and forces them to be re-generated. If any of t...

I am going to have to find some time to look at the actual code generated when you do this; since it seems as if DataStage is doing an implicit conversion to numeric for both sides of the conditional equation; which surprises me.

Orieos, When you create a hashed file with a very large modulus it takes the time to pre-allocate the disk space, and this time is reflected in the rows/second value. You need to have one job that creates/clears/deletes/re-creates the 2 hashed files, and and then another to do the actual writing for...

Unfortunately the u-umlaut is the representation seen for @FM (254) and that is the only character (apart from NULL) that cannot be contained in the key of a hashed file. In your case, unless you wish to work with NLS, you will need to re-map this character to some value that you don't use in your d...

Hello Madhu, welcome to DSXchange. The return codes that jobs bring back are documented in different places but offhand I would say that the BASIC handbook contains the most useful information pertaining to this question. The description of DSGetJobInfo with the mnemonic of DSJ.JOBSTATUS will show y...

Hello Vindy, a sequence job usually aborts when one of the jobs or other entities that it calls aborts. You said that you called the jobs within the sequnce individually - did they all run with no errors or warnings? What is the exact error message that your sequence aborts with? This almost always ...

Sengs, I've developed RTI jobs and would be more than happy to help you with urgent job creation. There are other consultants here who might do so as well, but their hourly rates are probably similar to mine. If your request isn't urgent and you feel like sharing what your problems are and what you'...

Using the RT_LOGnnn is certainly not going to address this. You are most likely more familiar with UNIX shell scripting than with writing DataStage routines, so the quickest and easiest approach would be to use the UNIX dsjob call for your job and to use one of the log command line options to return...

Elavenil, there are very many factors that affect performance; it is impossible to give a good answer to your question. You need to find out which part is making the performance decrease - it could be CPU, but it could just as well be connections to the database or factors which might seem completel...

Phil,

odd, I don't have a test env right now, but what happens if you do a

Over-reaction is not a problem - we all do that occasionally. But even with strings that contain only numeric characters the normal conditional operators will only cause an incorrect result if a conversion to numeric is forced; and that would only happen if the other operator is already internally r...

Phil, I've been using DataStage since version 0 of the product, and have been using the underlying UniVerse engine (actually doing product support for the manufacturers & consulting) since 1983. In all these years I have only come up with one or two cases where I was forced to use "compare()". S...

Append a semicolon (;) to your command. If you don't do this, then the engine will attempt to parse the command as a DataStage "SELECT" command, which has a different syntax.