DSXchange

Actually that routine works while testing . but was not implemented. There's another routine that can do the trick Arguments : UnixCmd, Seperator Command = UnixCmd; SepDef = If IsNull(Seperator) Then @FM Else Seperator ; Message = '' Call DSExecute("UNIX", Command, Output, SystemReturnCode...

This is a very simple routine executes a Unix command which is passed to it, and returns the output of the command. It Can be called from a transformer - but the command output needs to be formatted in the way you need ( watch out for field and item seperators, newlines etc) Input Arguments: Arg1: c...

Pretty easy if your data is sorted. You need at most a sort stage for unsorted data , with key change indicator pseudocolumn , us ethis in a transformer to concatenate strings if key col data is same, and use remove duplicates at the end with duplicate to retain = last .

Cool. It was one of my wishlist items (I guess for many others too) that a person be able to see the job log , but not start/stop/reset a job -The 'Operations Viewer' role goes a long way in addressing that gap. I had also heard of a DS Director lite client that was later discontinued.

Im wondering if the identifier '#SCTY_XREF_TEMP8' should have been #SCTY_XREF_TEMP8#' ?? just in case it was meant to be a job parameter

I do agree with the OCI stage. But this one has been biting me badly of late , so I used an ODBC stage attached to Oracle Wire protocol driver . Works Perfectly ! Update failures ( zero rows found) are captured with link variable REJECTCODE set to '17' Performance statistics or monitor reports 'thin...

Ideally issue a warning message in the job log . I remember Oracle Enterprise throws something like "Unable To use a row for update" ( though this error message has a differnt context) . I cannot insert these rows yet since not all columns ( some containg valuable info) are present in the ...

Im trying to update 7 million rows using an OCi stage. The job log says all rows were updated successfully , but I created another job and captured lookup failures from same table i was updating earlier - and i saw about 3000 rows failed the lookup - which means they did not exist in the table in th...

I just found out that while the number of records from the rowgen stage can be parametrized , the cycle limit cannot be - so I guess you are better off knowing the cycle limit (max for now and for all of eternity ) in advance :) . I had used this logic earlier to multiply num rows and get a counter ...

I havent worked with datastage 8.5 . However in datastage 8.7 I see a new role "DataStage And QualityStage Operations Viewer" exclusively meant for users restricted to the Operations Console Web application - Thse users will not need the datastage client installed to view job logs. Access ...

You can import a stored function using a stored procedure import. And Call it using a stored procedure stage. Works for us ( Oracle 11g ) . For convinience im capturing the return code of this function .

You can use a parallel job with transformer in v 8.1 and row gen and lookup for this as ArndW says - ONLY if the max possible value of input.col2 is known in advance.( Otherwise you need to determine it using an Unix Sort or andther datastage job - and set its max value as a parameter to the main pr...

Follow up : We've succesfully configured Operations Console . Our database is newly created on the metadata server. From the looks of it - I think is a really great new feature from v 8.7 onwards.

welcome. I hope the fix packs would cure the issue you are facing . And thanks for reminding me that job executables can also be migrated ..

Export all yiour jobs to a dsx file and search for the exact name of your dataset using Find. you will get a clue which jobs use it .