DSXchange

I also have Type 1 Dim with option of closing records that are not in the current load, but are present in the dimension table. I played a bit with SCD trying to set purpose codes on expiry date and current indicator fields, but Datastage requires having some Type 2 fields in such a case. This would...

I can't see a way to apply Data Elements anywhere in the STP stage .

Hmm, now I'm getting the following string in the file
"1600-2 - : : ". It seems just to format internal representation of the date, which is 16002.

I guess it is because of Date data type. Can't you use TIMESTAMP in Oracle? I know :wink: it's because of Date. No, I can't use Timestamp in Oracle. It is an existing SP and it's used by others in its current form. Try calling it a Timestamp(6) within DataStage. Tried it. Still getting the same err...

I have a stored procedure (Oracle) that has a DATE input parameter. The parameter represents the last process time and it must contain both date and time. I'm using STP stage to call this stored procedure and have a problem with passing full date and time portions. When I pass a Date (26,6) field to...

I have an Oracle connector stage with a Date(26,6) field that gets assigned with the result of to_date('20111023170825', 'yyyymmddHH24MISS'). Then in a transformer I convert it to VarChar(50) using the following Oconv(lnk_inp.D_DATA_STORE_DATE,'MTS') and store into the file. But the value in a file ...

I definitely need to be able to do checkpointed run. Thanks for the idea Craig. I might do as you suggest if there is a lot of noise ;-). In the setup we've got it should be relatively easy to do it, as there is one high level sequence for each stream that may require resetting (they are run by Cont...

I have a hierarchy of sequences in which failure of a sequence in the lowest level causes all the sequences "above" it abort (which is the expected behaviour). Currently, when the problem is fixed and I would like to restart processing from the beginning, I have to reset every single seque...

In some of my loads (via Oracle connector) I don;t want partial commits, i.e. I would like all the records to be commited or rolled back on job failure (this is how the current solution works using SQL loader). I understand that I have to set the transaction record count to 0. At the moment, the Arr...

Pandeesh, Did you manage to solve this issue? The same thing is happening to me and the routing can't properly locate the latest starting event for the job. This happens when I use the the syntax, you are using i.e. hJob = DSAttachJob(JobName, DSJ.ERRNONE) NewestStartId = DSGetNewestLogId(hJob, DSJ....

I managed to connect to the actual job's log even if the job is multiinstance and get all the relevant information. So it is all solved!!!

Unfortunately $ErrSource gives me the name of the activity within the sequencem, rather then the name of the job. But I have created a routine that interrogates the sequence log and extracts the actual name of the failed job. It then goes to the log of this job and extracts the last DSJ.LOGFATAL ent...

Sorry, but I don't quite follow. Do you mean having "a Routine activity to scan the server job's log." in the Exception Handler? But how do I know which log to scan? The error information is in the log of the job that failed. I suppose I would have first to scan the sequence log to get the...

Yes, passing user status via the direct trigger link worked. Thanks! It make the sequence design more "busy" though :(, but I might have to live with that. I have a more general question then. What is an example of the flow when using exception activity is the best option? In a case of my ...

The only other option I can think of is calling my routine as after-job and dumping the error message into the file and then reading it from the file when exception occurs. I would like to avoid use files though if there is another way.