DSXchange

The Any sequencer is a correct approach, which means you can use a single Notification Activity. Consider instead using an Exception Handler linked to a Notification activity (and anything else you want to do). The Exception Handler has activity variables that allow easy identification of the activi...

As a test, add ; echo $? to the command (perhaps in a separate sequence created for diagnostic purposes) to learn what the command's exit status is. Any non-zero exit status from a command will be treated as a failure.

Is dat_maj_socl referred to twice (or more)?

A Usage Analysis on a job will only find a "job" in which the selected job was used if the selected job is in a successfully compiled sequence. The Usage Anaysis will find the sequence. Usage Analysis is primarily intended to be used with table definitions, stage types, data elements, tran...

Is there a message in the status bar indicating that no items were found/selected?

You can't. This is the internal program used to initiate the run of DataStage jobs. It is IBM Intellectual Property.

Therefore you need to involve your official support provider. Most bugs in DSD.RUN involve them sending you a patch consisting of a new version of (the executable of) DSD.RUN.

Of course, the "someone" may have been a scheduled or background process, perhaps doing a backup or a bulk load or a reorganization or some such - it doesn't have to be a person.

This is usually a problem entirely within the client. Does it occur with the same jobs on someone else's client machine? The solution may be as simple as re-starting your client machine or, at the next step, reinstalling your client software.

WHAT is "completely dynamic", and what do you mean by that term?

We don't like to guess, it's so often a waste of time.

Even if you can do it (regen with the engine running) it is not a good idea. And you still need to re-start the engine services before the changes will be picked up anyway.

Only if your surrogate key is an integer and you're using a state file. The state file can be initialized from any source.

If you're using GUID for the keys, then you need to go back to the database using a query to generate these.

No.

DataStage lacks an RMM (read my mind) stage. You will have to tell it the file names. However, one job could be created with the file name set as a job parameter, if all three files have the same structure or you use the Schema File property (parameterize the schema file name too).

Dump the records' key values into a temporary table as well as into the real table. Then do a count query on the join of the two (based on primary key and therefore resolved rapidly in the indexes) and see whether it matches the count reported by DataStage.

What version of IE are you running?

I think Vincent's blogs will be the DSXchangers' best source of information.

(I don't believe in unnecessary duplication of effort.)