DSXchange

Hello all,

I re-created the job from scratch and everything seems ok
It's not clear what can have been wrong...

Thanks anyway!

Peter

Hello,

I just removed the warning trigger link from the job and still get a warning ???

Peter

Hello, Some more info: The jobactivity has one trigger with expression type: warning-conditional. This trigger leads to a routine that logs the name of the 'failing' job (using DSLogWarn). I also tried to create 2 triggers on the jobactivity (OK and otherwise), but this didnt work too. The compiler ...

Hello all, I have a sequence job in which a parallel job is called. Only if this job ends with warnings, the sequence job should trigger a warning. What confuses me is that the sequence jobactivity always triggers a warning, even though the called job has finished with status OK?! Any ideas what cou...

hi all,

I changed the execution mode of all stages up to the join stage to sequential and ... got the right output!!

It's not clear why parallelism fails here. Maybe you guys have some ideas about this.

Thanks anyway.
Peter

hi all, thanks for your advices. i create a testjob with a dataset containing 110 records and a dataset containing 5 records. I joined both datasets on a dummycolumn (both filled with 1) that i added to each dataset using transformers. The result: 550 records ! However, when I implemented the same s...

Hi all,

I have 2 datasets that I want to combine as follows:

Dataset 1:
col1
1
2

Dataset 2:
col2
1
2
3

result set:
col1 col2
1 1
1 2
1 3
2 1
2 2
2 3

I think the join/lookup stage is not suitable because you need to specify which keys in both sets you want to link.
How to do this?

hi all

- The mode of operation was already set to sort..
- How/where do I specify that the data is already sorted. Cant find the option in the aggregator?

Peter

hi

I added a sort stage and adjusted the partition type to same in the aggregator. Isn't that enough?

Peter

Sorting the data on the aggr key before aggregating performs no better...

Peter

The aggregation key is a varchar(11). The 300 output columns in the aggregator stage are integers, and decimals (most with precision: 15,3)

Peter

It doesn't seem to help (still 48 minutes), or may be I misunderstood your advise...

I added a sort stage before the aggregator stage to sort input on the aggregator grouping key. In both the sort and aggregator stage I selected hash partitioning.

Peter

hi all, my datastage job aggregates about 450000 records into 210 groups with 300 summed output columns. The outcome of this aggregation is fine but performance is very poor. Running the job takes 45 minutes to finish. I tried to sort the data on the grouping column before aggregating, but without s...