DSXchange

Parallel jobs do not like null values, especially in natural key fields, I suggest you default the NULLs to dummy values to make lookups, joins and comparisons easier. You never know what your NULLs are going to break: your current DataStage jobs, new DataStage jobs, database view joins, report filt...

I don't think MetaStage failed however it was never as popular as DataStage. Certainly analysts such as Forester and Gartner rated IBM-Ascential and Informatica more highly because they had metadata reporting products and if you put the time, money and effort into learning, installing, configuring a...

There is no specific Rochade metadata broker but you may be able to find a format they both support, such as UML. The readme file for MetaStage has metabrokers for UML 1.1 and UML 1.3. Rochade should have some type of UML export.

They post very good analysis of IT news. It would be beneficial if you go along to some of the events. I was invited to a couple events for the Australian branch but couldn't spare the time.

There is also my post Lee Scheffler interview - The ghost of DataStage past that describes the early history of DataStage starting before version 1.0!

Have a look at vmware . They let you run virtual machines on your laptop. You should be able to download a free vmware server and a free Linux virtual appliance already installed and configured from the VMware Virtual Appliances . You can run the DataStage server in Linux and the client software on ...

I uploaded the DataStage EE Remote DB2 Configuration guide (version 3) to IBM developerworks in this forum thread. It a good guide for configuring the DB2 Enterprise stage.

DataStage doesn't really know whether the load command worked or failed (in terms of number of rows loaded). You could write a DataStage Routine that opens the log file and parses it for success and failure message for use in log messages and Sequence job triggering.

There is no certification for server edition, only for parallel jobs version 7.5.x. You can find certification resources at My Squidoo lens or through searches on this site.

There is an option to take all the values in case of a Multiple look up "To specify that a link can legitimately return multiple rows: Select the link name from the Multiple rows returned from link drop-down list (note that only one reference link in a Lookup stage is allowed to return multiple row...

You could also import it as 11 columns and split the 11th and 12th columns in a transformer stage. Not completely sure if this will work!

You should also understand (and document in the job as an annotation) the business rule behind the date field. Do you set every row in a batch to have exactly the same creation date (which a stage variable gives you) or do you set a unique creation date for every row taken from the system clock (whi...

You can use the change capture stage to compare the source table to the target table and deliver just the changes (inserts, updates, deletes). You can truncate and reload the target table, if you use bulk load this should be quite fast. If you are doing this for a lot of tables then the Replication ...

There is not a lot of value in row-by-row ETL process if your source and target database is in the same RDBMS and you are moving the data and not really transforming it. As you say if it takes 6 hours from DataStage and 3 minutes in SQL Server then DataStage is simply not viable. I would be asking w...

You are posting in the parallel forum, if you are using parallel jobs you will need to go to the FAQ forum for the entry on creating a parallel counter within a Transformer. The routines you listed cannot be used in parallel jobs.