Seriously though consider your PC. You use it, it IPLs (reboots) this
irritates you because you had stuff open. And that's just YOU!
Now consider your IBM i server. It has 3, 5, 50 partitions on it. The
host partition gets some PTFs It IPLs. EVERY Guest goes to sleep, and
(hopefully) recovers perhaps an hour later. EVERY user on the machine is
stuck. And oh boy what if the PTFs fail? Now you're in a full blown
panic. Did the thing do backups before that automatic IPL? Likely not,
how would it know where to do that to? Can you recover? Where are you?
Out with friends? Had a couple beers? Out of town?
And what about all those other partitions? They need PTFs too, when will
they IPL? How much disk do all those PTFs take on all those Partitions?
What if they are small and now run out of disk?
Next consider the servers they communicate with, how many of those are
disrupted now? Were there scheduled tasks that didn't or couldn't run?
Did they wait or fail? What's the recovery?
Today's data center is rarely just one server and rarely can you state
"We don't do any work after 8 PM until 7AM" Almost every machine I work
on today supports web instances one way or another many directly some as
the DB Back end. You cannot simply bounce those things at will.
You MUST consider the environment. This includes timing, backups,
sequence, Staff availability, changes the PTFs may include, SLAs etc.
I repeat: NO!
- Larry "DrFranken" Bolhuis
www.Frankeni.com
www.iDevCloud.com - Personal Development IBM i timeshare service.
www.iInTheCloud.com - Commercial IBM i Cloud Hosting.
On 4/28/2016 12:05 PM, Rob Berendt wrote:
Larry,
We need details. Not that I suspect you prefer difficulty because that's
where your bread is buttered but I can see how others might.
Rob Berendt
midrange.com
