DSXchange

Not true. If you use direct Read and Write methods, RCP will cause entire rows (all columns) to be propagated even in the absence of a schema file. That's one of the reasons it can be so dangerous! What if the rows have 5000 columns?!!

Not if you don't have a licence or the right operating system for that piece.

Has NLS ever been enabled? Not "as far as I know" - please check. Execute the command $DSHOME/bin/analyze.shm -t0 | grep NLS to see whether it's currently enabled, and grep NLS $DSHOME/*uvconfig* to see whether it's ever been enabled.

If you create the hashed file in the account, the correct way to remove it is to use the DELETE.FILE command. Given that you have removed the operating system object, you now have two other tasks to perform to clean up; remove the operating system object for the dictionary portion and remove the VOC...

Because Windows has no native telnet service, DataStage provides a telnet service. Therefore you connect straight into a DataStage environment in the project that you select. If, from there, you need to get to the operating system, use the DOS command. If you want immediate return from a single DOS ...

I'd report the dead link to IBM if I were you. There are other paths you can take from the DeveloperWorks page; there is also a Search facility there.

Try LIST.DICT APM.BATCH to get the publicly-accessible column names to use in a query.

Beware that DESC is a reserved word in SQL, so would need to be encased in double-quotes if explicitly mentioned in a query.

The only formats that DataStage can use to import a table definition directly from a metadata file are Orchestrate record schema or COBOL file definition. All others (such as XML) require extra steps.

That has to be doable, but I don't have convenient access to QS at the moment. I don't imagine that rule set overrides are any differently done in Hawk than in 7.5. I would imagine it's a classification override, but I will need to research it. Maybe someone else will post an answer in the interim.

Unless your data have been through a Standardization stage the Match stage is likely to assign a negative (disagreement) weight to "Bob" and "Robert". But you could match on a phonetic of some kind - probably not one of the out-of-the-box ones, which require the first or final character to match. Be...

Beware that there may be other X characters in the key. In that case a different formulation will be required, such as

Is there any reason why performance would taper off after a certain period? Yes, the reason is that you're equating rows/sec with performance. Rows stop when all rows have been sent to the server, but there they're all queued waiting for the COMMIT. The clock keeps running while the database then i...

Configuration (creation/editing of configuration files) ought to be reserved for the DataStage administrator to do. All the developer needs to do is to select an appropriate configuration file for the job, by setting the value of APT_CONFIG_FILE (preferably as a job parameter).

Search the forum for the DS.REINDEX command.

Rows/sec is entirely misleading in this scenario. It's probably valid while the data file for the bulk loader is being written. However, the clock keeps running after that, so the rows/sec reported diminishes, while the bulk loader is doing its things. Do not react to rows/sec. It's a meaningless me...