IBM Mainframe Forum

I need two versions of same db2 pgm at same level EXECUTING SUCCESSFULLY. HOW IS THIS POSSIBLE

Example:



Version 1 - ProgramA – SQL QRY - SELECT COUNT(*) FROM TABLEA

Version2 – ProgramA - SQL QRY - SELECT COLM1 FROM TABLEA

Both the program versions ((with two different SQL’s) should be compiled, promoted and available at DEV level.

Can someone answer this with and without changeman

Seems like you'll need to pass the program some sort of parm to let it know which SQL to run.

Alternately, you could make a second ENTRY point. The name used to call the program would let you code which way it runs.

Or the sources could be saved to two different PDSs, the packages placed in two different collections, and the plans bound from those different collections.

Hello,

I need two versions of same db2 pgm at same level EXECUTING SUCCESSFULLY.

Why does someone believe this is needed. . .

Someone may have chosen such an implementation, but it is surely NOT needed.

Suggest you look into dynamic sql or simply have one program for each requirement. You wouldn't have to go very far with this "several of same" approach to reach an unmaintainable mess.

Is this something to actually be done or is this just some class experiment?

dick scherrer wrote:Hello,

I need two versions of same db2 pgm at same level EXECUTING SUCCESSFULLY.

Why does someone believe this is needed. . .

Someone may have chosen such an implementation, but it is surely NOT needed.

I assume that this is to be done in the development, not the production, environment (if it is in the production environment, I agree with you; nip this stupidity in the bud). At my site, we probably have dozens of load modules (each comprised of one to several thousand source modules, not to mention dynamically-loaded subprograms) with two or more versions (as many as half a dozen in some cases, I think) being developed simultaneously. Each development effort has its own libraries and DB2 objects (data bases, tables, package collections, etc.).

Akatsukami wrote:Or the sources could be saved to two different PDSs, the packages placed in two different collections, and the plans bound from those different collections.

just adding: If you have automatic package versioning (I believe it's default) the packages can be in the same collection, plan. because primary key of a package is (Location,collid,name,version)

yes this is for development n nt in prod

abhikulsh wrote:yes this is for development n nt in prod

Well, you have Mr. Goodman's suggestion as to how to do it, my suggestion, Mr. Scherrer's suggestion, and GuyC's suggestion. Pick one and implement it.