DSXchange

Thanks Craig

Craig,

Thanks for your quick reply. I was under the assumption that only hash partitioning in parallel jobs works by creating and storing the hash value.

More details would be very helpful.

Thanks,
mav

Hello Gurus,

When data is stored in Hashed files, is it indexed on key fields and stored?

Thanks

Thanks Ray

Hello Gurus, We have a Server job which has 1 source, 5 transformers (with a total of 30 lookups) and 1 target. The target is Oracle table which has currently 11 million records. There are 3 separate flows to the target: 1 for insert, 1 for update and 1 for delete. This job used to run at 250 rows/s...

Here's an useful update: 1) For few jobs target table had CLOB fields. Changed it to VARCHAR and reran the Sequence. No issues, ran fine. 8) 2) For other jobs which didn't have CLOB fields in target table, I found out that changing "Insert without clearing rows" to "Truncate and then ...

Here's an update:

When I run the Sequence (containing the Job) I run into that error but when I just run the Job it runs fine

Craig,

I'm wondering if it could be due to changes to the uvconfig file.

We made the following changes according to https://www-304.ibm.com/support/docview ... wg21430372

MFILES to 150
T30FILES to 300
GLTABSZ 200
RLTABSZ 200
MAXRLOCK to 199

Also changed values from shmtest output.

Thanks

Craig,

Thanks for your response.

DS Gurus,

We are again facing the same problem - Can't insert NULL into a field. If we Clear Status Flag for the Sequence and its Job and restart the Sequence it runs fine.

Has anyone run into this situation before? Any help is greatly appreciated.

Thanks
-Mav

Hello Gurus, We've been facing this strange issue recently: Our ETL jobs run every night and a minor subset of (random)jobs abend with Cannot Insert NULL into Column XYZ issue. Once we Clear The Status File for those jobs and rerun, they run fine. Just FYI - we recently migrated our code from Solari...

DSguru2B,

Thanks for your valuable suggestions

I'll give it a try and if it doesn't work I'll reopen this thread.

Thanks DSguru2B.

But why does the same job run faster on a slower machine?

How do I do that in a Server job?

Hello Gurus, We have a job as below DRS stage -> IPC stage -> Aggregator stage -> .... This job in a slower box A runs much faster than in a faster box B. In box B it starts off with a high rows/sec processing rate and after a few minutes reduces to 1 row/sec. The DRS stage has around 2 mil. records...