RE: Activation groups for beginners

[Carel Teijgeler <COTeijgeler@xxxxxxxxx>]

--
Just my contribution to this discussion (see below)

Regards,

Carel Teijgeler.

Activation Groups (AG), as I understand the concept. Correct me if I am wrong.

There are two groups of programme's (PGM): OPM and ILE. OPM PGM's are
written in RPG III or RPG IV. ILE PGM's are only written in RPG IV and uses
elements of the ILE concept.

RPG IV PGM's can be compiled with the following options:
DFTACTGRP       *YES (default) This AG is created at the start of the job
and I assume the name                          of the AG is the jobname.
                 *NO   all ILE PGM's using elements of the ILE concept
should be compiled this way;                       it requires the next
parameter.

ACTGRP  name (default QILE) a name provided by the developer, if the AG
does not                         exist at runtime yet, the system will
create it.
                 *CALLER PGM should run in the active AG: the AG in which
the PGM runs, that                            calls this PGM. Therefore
those PGM should run in the DAG, too.
                 *NEW the system will create a temporary new AG and assigns
a unique name to                       that AG; at the end of the call the
system will remove the AG.

In all cases: any AG has a name.

You can run a PGM in three kinds of AG's:

1) Default Activation Group (DAG). The DAG is the environment that runs a
mixture of OPM PGM's and ILE PGM's.

2) A Named Activation Group (NAM). In the NAM ILE PGM's will run that have
been compiled with the  name of an AG provided or with the special value
*CALLER.

3) A *NEW Activation Group.

Considerations in defining an AG to an ILE PGM are manyfold; think about
sharing resources, overrides, record locks, error handling, recursive calls
(call stack), performance.

For instance compiling an ILE PGM with ACTGRP:

*CALLER
This PGM will run in the same AG as it is called in (obvious); it can run
in all types of AG's. But if a job has started a number of AG's and this
PGM is called in different AG's, this PGM will be opened in all those AG's,
using the same resources. Sharing resources is not possible.

QILE (or any other name)
The AG will be created, if it does not exist. PGM's in different AG's can
call PGM's in this AG, hence the PGM is opened once.

*NEW
This AG is created and removed when it is required or not. This AG will
give the greatest performance overhead.

>Synopsis:
>
>1) Programs that run in the same AG
>    implicitly share resources.
>2) Shared resources include overrides,
>    commitment control, static storage,
>    Open Data Paths, SQL cursors
>    (there's more but these are the common ones.)
>3) There is no single "best" AG strategy.
>4) AG strategy is part of the overall application
>    design - see 1 above.
--