DSXchange

It's mandatory for quite a few stages, including Join. Aggregator is not one of them, however incorrect results will obtain if you aggregate on incorrectly partitioned data. A deliberately constructed bad example produced a report on four nodes showing all 200 states of the USA! As noted earlier, (A...

Only if the source data type is a string of some kind.

If the source data type is Date, then theoretically it can't be invalid unless it has more fractional seconds than are specified in your metadata, or is not constructed using the format specified in your metadata.

Oracle is... different. Nowhere else in the world has VARCHAR2 as a data type. When you import this into DataStage it has to be translated into VarChar, as used by the rest of the world. With NLS enabled and an NLS character set specified for Oracle, DataStage uses NVarChar, by which it means VarCha...

You will get metadata mismatch errors/warnings when/if you do that.

Is the user who is invoking from Datastage a member of sysadmin role?

Look at the "job starting" event in the log for JobB to determine with what parameter values its run was requested.

Use a Transformer stage plus a reject link from the Oracle connector stage. In the Oracle Connector stage select the reject link and specify under which circumstances rows are to be directed to the reject link and what other information is to be added.

Key-based partitioning IS needed in parallel jobs. A key-based partitioning algorithm will be selected under appropriate circumstances when (Auto) is left in place. This will always work but may be sub-optimal (for example other key-based algorithms than Hash exist). The nodes on which a Join stage ...

Only if your Transformer stage is executing in sequential mode. Otherwise the unqualified +1 approach is pretty much guaranteed to generate duplicates.

Depending on your target there are other strategies you can employ, such as a generated Surrogate Key or a database sequence.

Is there any reason you're not using an Oracle Connector stage? Do some comparitive tests for this also.

Please post the "job starting" event from the job log from an occasion when the job was invoked from a sequence. This event includes all parameter values received by the job.

The IN list in the WHERE clause is still being evaluated as INTEGER values, so there is no danger of truncation. The CAST is in the SELECT clause, which only affects what is "displayed" (returned). You should get one row back for each value matched.

Use CAST(RT.ID AS VARCHAR(10)) in the SELECT statement.

For job activity you don't get a UNIX exit status; you get a DataStage exit status (via the $JobStatus activity variable). Note in particular that, in a DataStage exit status, 0 does not mean "success" - it means "running". You're unlikely to see this downstream of a Job activity...

You have to count the matching characters. Assuming you have version 8.5 or later you can do this using the looping capability in the Transformer stage. Or you create a routine.