TSO exits are not designed to be dynamic but you can do things to make it work 
dynamically. I don't use these exits so I can't say specifically what will 
work. Here is what I would try. Be sure to consider overhead when implementing 
your exits.

1. Try placing the  the exit into LPA and use SETPROG to replace it as needed. 
Each user may need to logoff/logon to get the new version of the exit. 

2. If that does not work, then CSVDYNEX was designed specifically to implement 
/ call exits. This method would require you create a stub submit exit that 
simply calls CSVDYNEX REQUEST=CALL passing the R0/R1 that was received by the 
stub. This method has more overhead and could possibly be high overhead for 
very large jobs. To reduce the overhead, you could have use REQUEST=CALL at JOB 
(and first time) and have it return the address you actually want to call. Your 
stub exit would then save this address in the exit parms for subsequent calls 
and call this address from the stub.

There are other methods available but try these before you use them. Those 
methods were specifically disabled by IBM for a reason (e.g. security 
exposures). If you must use insecure methods, then only use them for 
development purposes on a standalone system.

Jon Perryman


>________________________________
> From: Dno <[email protected]>
>
>
>
>Can it be done? Install a new version, back off if necessary without an IPL?
>

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to