IBM Mainframe Forum

dick scherrer wrote:Howdy,

Thanks for the clarification - my dain brammage never got there . . .

So the "file" could have been read and the SQL generated into a temporary file, then run in a second step that executed these update queries (i.e. "batch" SPUFI). . .?

Yes. Although I believe that some of my peers are of the opinion that that doesn't create the necessary audit trail I don't claim that I've provided an adequate trail, but I've at least tried to acknowledge the need for one.

Many of the places i've supported do not allow "on the fly" changes to production data via QMF, SPUFI, etc.

They do support running canned generalized code that will perform these updates as well produce audit trails that are acceptable for that organization. These are scheduled as regular production and have "real" jobnames.