DSXchange

Kumar_s,

does this mean your COBOL PICture is something like:


What happens when you include a

to the beginning of the file?

Rumu,

what Naveen explained is true for Dynamic hash files, they are created at the OS level as a directory containing 2 visible files. Static hash files are created as a single file at the OS level.

Usually the slowest element is/are the outputs where you write; writing to a sequential file is usually very quick and writing to databases is usually slower than the reading part. What happens when you eliminate all your writes, does the hash file performance now remain constant? Is the hash file a...

Are you reading this hash file as the main input or as a reference? Usually a speed decrease happens when you write, not when you read. Please edit your job to make the output of this go to a sequential file or put a constraint in a transform stage right before your outputs that ensure that nothing ...

rumu, I recommend reading the actual manual, which you can download at Universe 9.6 Documentation The main difference is that the DYNAMIC type of hash file will dynamically adapt to file sizes - increasing and decreasing it's modulo as required and defined. This is excellent for files whose size you...

Kogcadoo, the MAXRLOCK affects you when doing SQL type actions on hash files - just as the description states, when you have more than this number of record locks within a SQL transaction it will go and get a file level lock instead and release the record level locks. I can't answer your second ques...

elavenil, check you job hash file definition carefully - if you use an account (the default option) then make sure the file name spelling is identical, it is case-sensitive. Same thing applies if you use a path - but then you need to make sure that your VOC file definition points to the same place. ...

The RLTABSZ also affects the maximum value of MAXRLOCK. I don't think that this value is causing your problems, since the locks help for the DSADMIN are normal DataStage database locks and not SQL locks. This is confirmed by the fact that your changes didn't affect the problem, I would recommend put...

Please tell us what the warning messages are.

(and YOU DON'T NEED TO SHOUT, we may be old geezers but can still hear well enough when the batteries are charged )

aolasoji_2, the DataStage Server has a function OCONV({YourString},'MCN') which will do exactly that. You do have the ability to use a DataStage BASIC transform stage in Px, but it will remove any speed advantages that the parallel paradigm normally has. So, if your data volumes are small enough the...

Hello Daniel, normally the concepts of "throw", "catch" and "rethrow" (see the common theme here ) are part of a process' error and exception handling mechanism. These handlers are common calls that should be included in standard libraries, but your transformers are aborting because they are not to ...

Craig,

I think that within the same Toolversion (currently I have "4" at 7.5.1) the jobs should be compatible. What does your .dsx say at 7.0?

If the internal revisions generated in the output .dsx file are the same (I don't have a 7.1 here anymore, but am fairly confident that they are identical) then you should be able to go back and forth at will within the dot-releases. It is only when the internal version numbers differ that you will ...

OK, I'm going to type this in and keep it short, so I am 100% certain that it won't compile: FUNCTION GetJobCategory(JobName) COMMON/DSJOBCommon/Initialized,DSJOBSFilePtr IF NOT(Initialized) THEN OPEN '','DS_JOBS' TO DSJOBSFilePtr ELSE CALL DSLogFatal('Unable to open DS_JOBS hash...

arindami,

I hadn't though of this before, but if nobody has an Ascential-standard method of doing this you could write your own snippet and use the (currently valid) fact that the category information is supplied in column 3 of the DS_JOBS file which is keyed to the job name.