DSXchange

Set Record Delimiter String property to "DOS style". This will correctly invalidate the Record Delimiter property.

Perhaps a range lookup where the range is between 0 and source value?

You do not need to run Sort in Sequential mode. What you do need is to specify "Sort Merge" as the collection algorithm on the input Partitioning tab of the Sequential File stage. This will collect the rows from all nodes preserving their overall sort order.

While the job is running in parallel (let's assume on four nodes) you will have four counts. Nothing will sum those together unless you create more stages to do so, or unless you run your Aggregator on a single node (that is, in Sequential mode).

Use a Copy or Modify stage to change the name of the join key column on one of the input links so that it matches the name used on the other input link.

Neither. It does a count on each node. End of story.

The reference data for a Lookup stage is loaded into memory.

Which rows are loaded to which node depends on the partitionong algorithm; Entire places all rows on every node. Key-based algorithms (Hash, for example), places a discrete subset of rows on each node.

You need one DSN per workbook.

Each worksheet is described as a (system) table using the ODBC driver for Microsoft Excel.

Verify how character storage is encoded into your Teradata table (perhaps via a SHOW TABLE command) and match that in your DataStage NLS map setting. Probably one of ISO8859-1 or UTF-8 will fill the bill.

The DISPLAY environment variable is used to specfy which display to use. You can find more about it using an internet search.

You need to involve your Teradata DBA and your network administrator - clearly the problem is an intermittent network connection failing to be made.

If you use Failure triggers, you can bring them together using a Sequencer. If you don't use Failure triggers, the activity variables for each job activity continue to be available downstream, so you can combine them with a logical expression (for example in a nested condition activity) ahead of gen...

You need a UNIX user. dsadm is the conventional name, but there is no restriction on what the name is. This user and its primary group (conventionally dstage) should be the owner of all DataStage project objects, as well as most of the Engine objects (but be careful there - there are scripts to help...

There are some really good examples in the manual, which are also available on-line from the IBM Information Center.

Check the write mode property - you should have Delete then Insert as one of the options.