DSXchange

Hello there, Recently I recognized that a job parameter name, that I expect to be case sensitive, seems to work correctly when being used in a SQL database stage (in my case it is Oracle) in a SQL statement. It seem to be recognized at run time correctly even if being used in different upper/lower l...

At least in DataStage 11.x there is a function called TimestampOffsetByComponents, which also supports microseconds to be added/subtracted in the seconds-component... :D

Cheers
Klaus

Tony, "The first solution I tried was allowing multiple lines as output - which is the universal solution that I seem to come across everywhere, including on some of the threads here - however I cannot use that because as we know, activating that option throws out more lines than in my source t...

Not sure if this is what he did, but as per my experience you have to add a clause like {text, width=15} to get it to work.

i.e.: mydecfield:decimal[15,0] {text, width=15};

Best regards
Klaus

Are you working with a GeoDB or extensions? These usually have functions to calculate all sorts of geometrics.

Klaus

Thanks Arnd,

I thought at least you must have seen this...
We will try this way. I is really annoying...

Yes, long time no see - are you back to Germany?

Best regards
Klaus

Hi there, has someone ever experienced that a dsx-export may contain numerous (thousands!!) of backslashes - one after the other. This seems to happen with server jobs that contain an ORACLE connector stage only. I cannot verify this exactly, but most of our jobs are parallel jobs and the server job...

The solution is: It has to be a "sparse" lookup in PX. And the lookup SQL then looks e.g. like this: "SELECT ORCHESTRATE.WOTEXT10 TEXT, REGEXP_SUBSTR(REGEXP_REPLACE(UPPER(ORCHESTRATE.WOTEXT10), '[^a-zA-Z0-9]', '', 1, 0), '[BK][0-9]{1,8}') TRANSAKTIONSCODE FROM dual" ORCHESTRATE.W...

Hi folks, I am in the process of redesigning server jobs to parallel. I have a lookup to an ORA-DB in the old server job that has is user defined sql: "SELECT :1 TEXT, REGEXP_SUBSTR(REGEXP_REPLACE(UPPER(:1), '[^a-zA-Z0-9]', '', 1, 0), '[BK][0-9]{1,8}') TRANSAKTIONSCODE FROM dual" and runs ...

Hi MT,

thanks for your suggestion. As I mentioned in my issue describtion, leaving "Instance" blank gives an error. As I mentioned I am on windows, so there is no dsenv....

Best regards
Klaus

Hi there :D I have an issue with DS(IS) 9.1 and DB2 10.1, both on one machine, OS Win Server 2008: I can run "classic" server jobs with DB2 API stage without problems. When trying a parallel job and DB2 connector stage I get "An error occurred setting DB2 instance to value . The activ...

Hahaha... this bug is still there in 9.1(!).

Klaus

I faced this error once having a corrupted dataset afterwards. I was able to delete the dataset with the dataset manager. Having done so the job then ran fine creating a new proper dataset.

Klaus

Hi Sachin, we do this running DSJobReport (report type 2/XML) for every job as an after job subroutine. It delivers everything you mentioned exept a Batch_id, which is your own thing. Later we process these XMLs to be stored into a DB table for auditing. While this works pretty fine you have to be a...

Did you try "wait for file" event in job sequencer...?

Klaus