Re: MONITOR performance

["Buck" <buck.calabro@xxxxxxxxxxxx>]

> Have a need to catch decimal data errors so that the program can handle it
without
> crashing.  When I suggested using MONITOR and trapping errors that way, a
colleague
> offered that the *PSSR might be a better performer.  I'd like to avoid
PSSR because
> MONITOR seems to offer better recovery options, so that processing can
continue.
> The program in question is a RCVJRNE exit program.
>
> This is an error that we expect will nearly never happen, but if it does,
crashing
> the program is "not an option" the client can live with.  The error did
happen in
> testing yesterday under circumstances that we don't expect to see in
production, but
> that we can't rule out either.
>
> That said, the solution should provide the same performance as if a DDE
never
> occurs.  I.e., the MONITOR group or the *PSSR should never impact
performance when
> there is no DDE.

Neither MONITOR nor the *PSSR have an impact on the code when an exception
does not occur.  There is possible performance impact related to the program
status data structure and fields beyond column 80.

In brief, MONITOR is better because it is more granular.  It will let you
handle the exception when you get it.  The *PSSR takes you out of the
program logic into a separate subroutine, and makes it very difficult to
continue processing at a reasonable spot in the logic.  I use them both.
*PSSR is my 'last ditch, no white messages ever' exception trap.  In my
*PSSR, I issue a DUMP and then leave the program.
  --buck