DSXchange

You'll need to test to verify, but from other stages I've used you get one connection per stage regardless of the number of output links.

Welcome aboard! :D That's all there and still intact, AFAIK, so shouldn't be an issue. Sequence jobs still exists, DataStage BASIC is still there so unless you are doing something 'tricky' based on an intimate understanding of the repository structures I would think you'd be fine. Hopefully you've s...

Magesh_bala wrote:What are different types of transformations available in Datastage..

This smells way too much like an interview question, something we're not here to answer.

If it isn't, please be more specific as to what information you are looking for. A lot more specific.

The 'warning' is simply the fact that your update sql did not update any rows. You eliminate it by not sending updates for non-existent rows. Use normal ETL techniques - existing keys in a hashed file for instance - to pre-qualify your work, inserts versus updates.

What 'remote source'? If there is a native stage, use it, otherwise you'll need to create something and use ODBC.

You'll probably need to fall back on Oracle tools for this, depending on the nature of the 'transforms' involved, the frequency, volume, etc. - something you didn't mention. I was hoping that by mentioning the fact that you didn't mention certain things that the mention of them would get you to... ...

Really?

Huh... gonna have to look that puppy up.

In case anyone wants to see the beginning of this conversation, check here so we don't repeat advice. So, you are saying that you were able to run this job successfully by pointing it to a test instance? If so, that would imply that, yes indeedy, you have an Oracle client installed on your DataStage...

Ah, I see... this is probably why they started over in the Server forum.

Yuck, a LONG ? Bad choice of datatype. There's not much you can do with a LONG under any circumstances, let alone using an ETL tool - there's too many restrictions. You'll probably need to fall back on Oracle tools for this, depending on the nature of the 'transforms' involved, the frequency, volume...

Your 'bad' record wasn't 'rolled back', it was never accepted by the database. Rolled back imples it went into the database and then some other error cause all current 'good' but uncomitted work to be undone. How do you make it abort? If you were using a native stage rather than ODBC, there's usuall...

Been there, done that, answered the question already. Curious how many incorrect places this question will get posted?

That's ok, but you should have expanded on what you are doing. What did you set the DRS stage to - Oracle? ODBC? Do you have an Oracle Client installed on your DataStage server? If so, have you configured DataStage for Oracle? Or has someone?

In that routine i am generating Job Report and then fetching out the Job Status of the current Job and sending this Job status to TIVOLI. Again, help me out here because I don't know beans about Tivoli. I've asked this before and you don't answer - how exactly are you 'sending this job status' to T...

Something is not as you've stated - if the job finishes in an Aborted status, any uncomitted rows will be rolled back. So either it isn't really aborting or you are actually comitting the rows. You need to help us figure out what is actually going on.