DSXchange

I have to develop a single generic job for all the tables to populate into target. Why? Last time I did this exercise (837 tables) I needed four separate generic jobs, because of slight differences (for example some tables had BLOB data). If you need to effect a transformation you need to name the ...

I suspect the database activity logs would be the best place to look. Ask your DBA.

Surrogate Key Generator stage wants an unsigned long integer. This is probably why it's complaining about -1 being invalid.

Array Size and Record Count are related. With either property selected click on Help.

Yes.

The method by which you access Passport Advantage and establish your entitlement has not changed.

Welcome aboard. This behaviour is governed primarily by the transaction record count specified in your Oracle connector. Each time that many rows is sent to Oracle, DataStage also sends a COMMIT. You can change the behaviour to "all or nothing" by setting the transaction record count to 0 ...

I still don't understand the requirement completely. Can you spell it out in words? Could it be as simple as a HAVING COUNT(*) > 1 clause in an SQL statement?

Almost certainly. When was the last time you re-indexed this project?

Do you get a different result when running the job from dsjob versus running the job from a DataStage client?

I envisaged something like Shane's solution, but abandoned it because I assumed that your possible range of headings was infinite. Solutions such as Shane's are fine for a limited, finite, set of headings only.

Do you mean Column Import stage? There's no such thing as an Import stage.

One biggie is the fact that only HTTPS is supported for various kinds of connections, including connecting clients. Once you've nailed the new URLs (with new port numbers and HTTPS as the protocol), you'll probably need to update your browser Favorites too! You should set up key files (for https) fo...

So you have deleted rows from the target table even though their keys were not in your source data?

In that case you could have specified "Truncate" as the table mode. That would probably be faster.

UAUITSBI wrote:The column names are not fixed they vary dynamically.

How do you envisage "some sort of sequence number" would help?

One of my clients has asked this intriguing question. Is it possible to build up to an active-active high-availability configuration in four phases? The metadata repository is Oracle RAC 11g. Phase 1 - Stand-alone machines for WAS, MDM, IIS Engine ("the three"). Phase 2 - Like phase 1 but ...