DSXchange

The main tool for conversion is located between the keyboard and the chair. Analyze the logic of the server job, identify need for parallelism (and for partitioning and sorting), then design a parallel job from scratch to re-create the logic of the server job. Forget everything you know about server...

Why only one node in the DB2 node pool? Surely this will be a bottleneck. Create extra nodes in this node pool on the Db2Server01 machine. Unless you do this, all DB2 operations will be sequential.

Have you imported the metadata (table definition) for this file into DataStage so that you can specify the column widths and the fact that it is a fixed width file from the outset?

The WHERE clause can use the LIKE operator followed by a regular expression. This is documented early in the chapter on the Filter stage in the Parallel Job Developers Guide manual.

Orchestrate is the name associated with the parallel execution engine that Ascential acquired when they bought Torrent Systems. Initially Ascential called this product the "parallel extender" for DataStage, which is the source of the "PX" - this nomenclature is now obsolete; it was a decision made b...

Have you taken the IBM class DX437 (Administering DataStage) or some equivalent? This provides you with the skills you need.

IBM do have some tools, but you have to buy the service and agree to your jobs being "off-shored". I support Mike's comment totally.

U can not help, because U is involved in a server-only project. Creating a reject link from a stage type that can have an arbitrary number of output links (such as Filter, Switch or Transformer) requires a slight variation in technique. What you have to do is to draw the extra output link as a regul...

That would depend entirely on what firewall software you are running. There are lots of these. Consult the associated help or documentation for your specific cases.

No.

There is a server-side export utility (for the first time) in version 8.1.

The dsjob command may have -export and -import options, depending on precisely which version of DataStage you have.

I happen to know that U is currently involved in a server-only project, so that your advice is not appropriate.

Merge stage requires de-duplicated data in the "update" input(s), and in the "master" input if there are more than one update inputs. Merge stage requires that all inputs be sorted on the Key column(s). Getting the right results also requires that all inputs be partitioned on the Key column(s).

A. UNIX will not permit two simultaneous writers to the same file.

There's no such thing as an internal format of a timestamp in server edition.