|
JES2 Exits
|
We would be very neglectful if on this page we didn't issue a warning about JES exits. JES
exits have the potential to make your operating system useless so
extreme caution should be exercised before implementing new JES exits
or recoding old JES exits. If possible always use a sandbox system for the testing of your
JES exits and use dynamic activation and deactivation of the exits on a quiesced system.
|
|
|
Mellon Bank Modifications
|
If you are looking for, what used to be called, the Mellon Bank Modifications
and some supplemental information then follow this link
|
|
|
Activating/Deactivating
|
Many (but not all) of the exits can be activated or deactivated 'inflight' and
there are basically two ways of activating and deactivating a JES exit, only one of these is a permanent activation (across JES2 restarts).
Exits
that create token pairs can not be temporarily deactivated,
these should be placed in (or removed from) the JES2 initialization
deck and JES2 restarted in order to activate or deactivate them.
It is strongly recommended that any changes to these token pair
exits are changed via an IPL as the tokens can survive a JES2 hot
restart.
Permanent Activation
Permanent activation or deactivation is achieved by way of adding or altering the
LOADMOD and EXIT statements in the JES2 Initialization deck and then IPLing the system. The LOADMOD and EXIT(nn) statements must come before any new statements using any keywords supported in the exits themselves. This is why the recommendation is made that new initialization statements are placed at the very bottom of the initialization deck.
The
format of the LOADMOD and EXIT(nn) statements is documented in the IBM
documentation and will not be covered in this document other than in
the portion of the document that describes the current JES2
Initialization implementation for all of the exits described in this
document.
Temporary Activation
Temporary
activation or deactivation is achieved by using the $TEXIT and the $ADD or $DEL JES2 commands.
These commands will only work if the JES2
subsystem is already active. These commands are ineffective if the exits themselves are already called.
In other words if an exit goes into a loop issuing the command to deactivate the exit will not break
out of the loop for you. Under these conditions an IPL is usually the
only way out of the loop.
In order to deactivate and remove an exit issue the following commands:
| |
|
|
$TEXIT(nn),ROUTINES=(-EXITnnC)
$DEL LOADMOD(BSLJXnnB)
|
|
|
To activate (or reactivate) use:
|
|
|
|
$ADD LOADMOD(BSLJXnnB)
$TEXIT(nn),ROUTINES=(+EXITnnC)
|
|
|
Note:
It should be noted that there are several exits that require a slightly
different $ADD command as they reside in CSA and so the $ADD command
for these would be:
|
|
|
|
$ADD LOADMOD(BSLJX50B),STORAGE=CSA
|
|