DSXchange

Tunde If this is one row then you can call the same function 3 times or five times with 5 different columns. If you need 5 lookups to the same hash file or table then create 5 links and join on a different column to the same table. You do not need a shared container. If this is all one row then you ...

its really the same one row spread out to perform lookups on some columns, then colapsing back to be one. row -1,2,3,4 trans 2 and 3 ---> shared container 2 = a, 3 = b ---> output 1, a, b, 4 im performing a lookup funtion on the columns for conversion and then merge the results back to the original ...

Isabelle You are right. I was thinking about this file as input not output. Rays idea will get you close but you will lose the last record. I think this is cleaner. You will always output a detail record. Do like Ray said on the separator character. You will need something like this because you cann...

Its not possible directly. Before/after invokes a Routine. You could write a Routine (of type before/after subroutine) that uses BCI functions to connect to your data source and execute any SQL, but would that be the same session? ----- Original Message ----- From: gxganes@regence.com Date: Mon, 7 J...

It certainly can be done. Since the target is a text file, probably the easiest way is to output a single column made up of the header and all the detail lines, separated by line separator characters (LF on UNIX, CRLF on DOS), and ensuring that the separator character on the Format tab of the Sequen...

Kim, Im pulling data from a database (joining a header and a detail table) and then will deliver the data in a file to a customer. So Ill be pulling from two tables on the source side and creating one target sequential file containing multiple sets of header and detail records. I think you were goin...

Its hard to tell, exactly, whether you are trying to avoid the 5 calls to the lookup, or looking for ways to streamline the job design.... but it would seem that a pivot type functionality is what you are mostly looking for, to ease the design....., and then a sort of "reverse pivot" to bring the da...

Hi all, I have a simple problem. its a little hard to explain but I will try. im on Datastage 6.0 on solaris. I have a row to process. Out of this row, i have five columns that needs to go thru a conversion in the same lookup function. in order for me not to have to call this lookup so many times, i...

Isabelle You need to read each line as one big field if there are different number of fields in the header records versus the detail records. You will need multiple output links if the header records update different tables than the detail records. Setup stage variables to store the last account num...

Hi, I would need to issue the following command from datastage OCI8 stage. Is this possible to do? I see a before and after tab page but do not know if an ALTER session can be issued. ALTER SESSION SET SKIP_UNUSABLE_INDEXES=TRUE The reason we need this we do parallel inserts using multiple datastage...

Hi, yes!! You can perform ALTER SESSION statement in the before/after tab. Regards, Riccardo ----- Original Message ----- From: To: Cc: Sent: Monday, July 07, 2003 6:48 PM Subject: Using ALTER session from datastage > > Hi, > > I would need to issue the following command from datastage OCI8 stage. I...

Hello, I could use some advice. I need to create a file that has header records and detail records. I have some transactions on accounts. So for each account I need to create a header record and follow it with the transaction records for that account. There are to be multiple accounts within one fil...

Dont know which version of DataStage you are using, but recent versions of the Job Sequence GUI should offer you "UserStatus" as one of the Expression Types on the triggers tab of a Job Activity. So if you can make your job set a user status (as you suggest below) of 1 or 2, you should be able to ju...

That would do it. -----Original Message----- From: tom.peter.nielsen@nordea.com [mailto:tom.peter.nielsen@nordea.com] Sent: Monday, 7 July 2003 6:36 PM To: datastage-users@oliver.com Cc: datastage-users@oliver.com Subject: RE: Job return codes Yes, but i didnt see how to use it from the sequencer. I...

Yes, but i didnt see how to use it from the sequencer. I guess that the correct way will be like this: Sequencer: First run a job that uses DSSetUserStatus Then from the sequencer, run a routines that uses DSGetJobInfo to read the user status, and sets the return value to the coresponding value. Hav...