DSXchange

I think that planning now for version 8 shows eminent sense; it is such a hugely different architecture. If you can get your hands on the Hawk presentations from last year's IBM conference in Las Vegas, or this year's in Anaheim, you will be able to plan better. Alas, there is little else concrete o...

It can be either, it can also be a non-first-normal-form (NFNF or NF2) database. It doesn't matter; it does the job. As implemented for DataStage Repository, the best description is a non-first-normal-form database. The DataStage Engine database ceased to be called UniVerse with effect from version ...

The MatchField() function will work in this case.
MatchField(InLink.TheString, "0N0X0N",1) for the first numeric field, MatchField(InLink.TheString, "0N0X0N",3) for the second numeric field.

Read about this function in online help or in the DataStage BASIC manual.

Outline your job design. Enclose it in Code tags and use Preview to get it to be a legible picture. In particular specify what stage types you have used.

RT_LOGnnn tables are not indexed, so there is no gain trying to reindex them. Most problems with logs have to do with them becoming too large (2GB) or becoming corrupted by other means. Although technically not the best solution, a quick fix is usually to clear the log file. Errors writing to path u...

For business rule changes the correct response is to preserve things exactly as they were, and augment your process metadata with information about when and what changes were applied to the business rules. It is a fundamental tenet in DW that we do not rewrite history. In the case of a bug it's a di...

Both will be in DSX file, but only the most recent (the appended one) will appear in the selection box.

The only way to retrieve the earlier one is to cut the DSX file.

Welcome aboard. :D

You are attempting to implement what is known as a "slowly changing dimension". A search for this term, and/or for "SCD", will yield discussion and techniques for implementing same in DataStage.

Of course it's asking for key - it wants to know in which row(s) to perform the update. If you want to update every row, fake it - specify any column as "key" - it ends up in the WHERE clause - and a condition that is always true (for example WHERE key = key). You may, depending upon the stage type,...

WHY does client only want Packager Wizard?

It is outdated and superseded technology, at least for job deployment. And if they don't understand "released jobs", then they are speaking from a position of total ignorance.

Challenge them!

How much memory do you have? Have you monitored memory to determine whether you really are running out of memory during this import process?

A hashed file is merely a mechanism for implementing a database table. Therefore you can search with a regular query. What is the nature of your search, and do you want to do it within DataStage or externally? Do you want a non-key lookup, or some other kind of search (for example "does this value e...

At the risk of being less efficient, you can use an Execute Command activity to run a script to execute SQL against a database. You may need to strip extraneous newline and white space characters from the command output.

Have you reported this to the Hawk beta team? Only DataStage TX is actually available in version 8.0 - all the other components are still in beta testing for this release. The fact that 7.5.1 can not properly load an 8.0 library is not a total surprise.

That IS a single statement.