DSXchange

All of a minute ago!

STOP thinking in C programming terms. There is NO zero element here. It's a string! It's a string delimited by space characters. ScriptPath = Field(InputArg, " ", 1, 1) DatabaseName = Field(InputArg, " ", 2, 1) UserID = Field(InputArg, " ", 3, 1)...

This is not C. InputArg is not an array. It is a string.

Therefore InputArg(0) is not a valid construct (unless there is a DIMENSION or COMMON declaration that you've hidden from us).

Please post the detail of what gets logged by the first two calls to DSLogInfo() in RoutineB.

Moderator: please move to parallel forum

Please do not post in the wrong forum; it's annoying and wastes time trying to get it right. Server job answers to parallel job questions don't help anyone.

That's the generated OSH, which does not show partitioning or, indeed, anything about parallelism because the latter is not determined until run time.

I asked for you to dump and post the score.

This is something you will need over and over: the ability to generate the current date/time as a timestamp. I advise that you create a Transform or even a Routine. I call mine NowAsTimestamp. Oconv(Date(),"D-YMD[4,2,2]") : " " : Oconv(Time(),"MTS...

Please dump and post the job score.

Welcome aboard.

The answer is: no-one can help because it is not possible to do it through DataStage server.

Sorted data, stage variables. Search for the rest of the story, how stage variable may be used to "remember" a value from the previous row.

The 20MB per node in the Sort stage is memory, not scratch disk. The process will go to scratch disk if that amount of real memory can not be provided on demand. Assuming nothing else is happening, you need 216 million x 20 bytes (approximately 420 MB) of scratch disk space to handle the worst case ...

Even minimal research on your part would have discovered it.

You had a filesystem related error. Nothing we here can do about that.

Hi, The job flow is as follows: Input ---> Transformer ---> File---> Table In Transformer, Suppose if a column address needs to be divided into address1 and address2, the derivation for that column(address) is given as Ereplace(FromProgressICSP.addr, ";", Char(13):Char(10), -1). Please help me how ...

Maybe they want to replicate the Repository in a UniData database?

Maybe your super-enhanced caffeine.dll is interfering with proper execution of vmdsedit.dll.