DSXchange

If you still have the .dsx file (i.e the exported file containing the routines)... then search for ReadOnly phrase... . . OLEType "CRoutine" Readonly "0" Category "EPM89_Routines\\Generic" . . Change the value "0" to "1" for ReadOnly prameter... and ...

I think.. when you rename the input link.. the key expression for the first field changes..but for the remaining fields..it does not... rite ? I also experience this in DataStage 7x , Server, Windows... Its a know bug from Ascential.. I will post the link of the post I found on the forum for this......

While writing the Hashed file use TRIM function to remove any extra spaces from left and right.. atleast for the key fields.
While writing the Key Expression, use TRIM function in the same manner... so your key expression would be... HTH

I agree with Ray's suggestion...

Analyse the cause of DataStage warnings which you are receiving now... Handle them... and if later any warning occurss then it would be a genuine error.

Time to mark the topic as Resolved please

Possibly..what Ray is suggesting is ..to dump a NULL into the seq file...anyhow you mentioned that the file is being deleted at the end ...so whats the worry ?

Ray has pointed out correctly that the command stage executes a Command not a DataStage routine... Also, I wonder if we can have a design, where there is an Input Stage, a transformer , BUT no output !! (This, I am saying, becuase the seq file in your current design is being considered as Superfluou...

"Array Size" is the number of rows to be processed at the same time... "Transaction size" is the number of rows after whicha COMMIT needs to be issued...

Ideally..Transaction Size a multiple of Array Size.

Also, Once a value , fetched from the SQL server 2k5, has already been stored in a hashed file... you can use UtilityHashLookup to retrieve that value again and again....

I have one approach for this: 1. Job1: Store the value fetched from SQL server 2k5, into a hashed file, keyed by single key and some constant value, say "Dummy". 2. Routine Stage - R1: R1 will use UtilityHashLookup routine, to fetch the value back from the Hashedfile (here provide key valu...

Like this approach too.... Fantastic !!

I agree with ArndW's suggestion...using TRANSACTION_SIZE of 0 is the best option for rollbacks... so if the job aborts due to any reason...then the data will not be committed. Also, I would suggest to load the child table first..and only when it is successfully loaded, proceed to load the parent tab...

Also, try running the job with a different input data set...

And , inside transformer ? What processing is happening on the data fetched, before being written to the seq file ?

.... I have to set a value selected from the database into the user status. I retrieve the value (a single value) from an odbc stage, via a query on sql server 2k5. I need to make this value available to all the sequencer this job belong to ...... ...all the sequencer... or ... all the sequences .....