We have had some spirited internal discussions on SMP/E best practice for 
applying z/OS maintenance while maintaining an environment for emergency PTFs 
as neededI upgrade z/OS version on a two year cycleCurrent process for a lot of 
maintenance with about a year's worth of RSUs        
   - ACCEPT after about a year of a stable z/OS
   - create a new target and dlib CSI by copying and change relevant 
definitions of existing CSIs
   - upload the GIMXSID output, order the latest RSU (recommended on a 
quarter-end RSU), APPLY to new target CSI pointing to new RES libraries
   - IPL and test the new RES in the sandbox LPAR
   - incrementally roll out to development, production, then customer LPARs
This allows an emergency/unexpected PTF(s) to be APPLYd to the original RES 
libraries in the middle of the updated rollout
The discussion is - is the ACCEPT necessary as a baseline if you are creating a 
new target and dlib CSI with its own associated libraries? If unneeded, that is 
a significant labor savings
In the past and infrequently, we encountered PREREQ/COREQ chains that required 
a lot of RESTOREs, and doing the ACCEPT first avoided that as we remember. This 
ACCEPT process has been successful in my 13 years at my present company - never 
gotten in the ditch
Another proposed alternative is duplicate all SMP/E libraries and zones and  
customize, with each maintenance version having its own discrete SMP/E 
environment, but isn't that the point of one global with multiple target and 
dlib zones?
By our SMP/E configuration, I realize REJECT and ACCEPT will delete the PTF 
from the PTS making PTFs unavailable to a global zone with multiple target 
zones, and the final fallback is always restore the backup and start over
I reviewed Share and IBM-main, and didn't see any presentations or discussions
Thoughts and best practices, anyone?Just looking for an overview
Opinion part:Yes, secondary discussions about shorter maintenance intervals is 
important, but we don't have the labor or IPL schedule to realistically shorten 
that interval
Will being more current on maintenance become an issue?I don't know about your 
shops, but I have never seen such a high rate of change/new products in my (oy 
vey) 49 years as a sysprog, and current maintenance will assuredly be an issue 
soon with some unscheduled PTFs being APPLYd already
ThanksDavid



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

Reply via email to