Re: "Pretty" announcement for End Of Support for Power 9.

Patrik,
This is one of those cases in which you fervently believe in what you
believe in. I've explained what I can and to try to do any more is a waste
of my time and a waste of yours.
I will grant you that "And the heap of electronic waste is becoming bigger
and bigger…" is actually a concern I've heard elsewhere.

I will say that getting an IBM repair technician to come on site and
replace an internal drive that's hosted by vios to IBM i is quite an
experience. As soon as they ask "What operating system does the VIOS
host?" you know you haven't got a snowball's chance of having a good day.
What was going to happen is they are going to send someone out who, when
placed at the VIOS command line the first thing they were going to type in
is STRSST. Then you really know you are screwed (STRSST is not a VIOS
command). I've been through this more than once.
I say "WAS going to happen". Because now they make it clear in no
uncertain terms that it is now the customers responsibility to know how to
break and rebuild that vios mirroring and all the other steps involved.
The sole responsibility for the IBM technician is to pull the old drive out
and put the new one in. Any data loss due to the CUSTOMER not doing the
mirroring right, quiescing other operations, etc is not on them. It's in
writing.
Oh, and if you are behind on VIOS you will be hazed like a (raw
recruit/college freshman/new (whatever))

Speaking of VIOS, do they offer Extended Service on unsupported versions of
VIOS?

On Tue, Feb 4, 2025 at 1:04 PM Patrik Schindler <poc@xxxxxxxxxx> wrote:

Hello Rob,

sorry, can't resist…


Am 04.02.2025 um 18:22 schrieb Rob Berendt <robertowenberendt@xxxxxxxxx>:

Many people are under the belief that EOS is simply a scare tactic to

get customers to buy new hardware.

I'm one of those. Although I'm aware that there can be exceptions. Most
often it's just arbitrarily declaring EOS to generate new opportunities to
sell new things.

1 - Service is based on parts availability. Oh, you lost hard drive

model xyz? Sorry, we no longer have that in stock.

Uhm. I presume, SSDs and NVMes are the same for Power 9 and 10?

Same goes e. g. for PSUs. When a vendor constantly changes mechanical
attributes for his PSUs for slightly newer hardware, it smells. Dunno if
this is the case for Power6, 7, 8, 9, 10.

2 - New CVEs about security vulnerability will not be addressed.

Normally these are fixed with Firmware updates. While one might think
they'd all be discovered by now (again thinking it means whatever they want
it to mean) there has been recent updates in the firmware for just such
things.

Software always has bugs, it's simple statistics. IBM surely is able to
fix security holes in central places, and have their shiny Ansible workflow
build firmware blobs for Power 9 and 10.

But this very topic is a the heart of "scare tactics". Why declare
perfectly working hardware being obsolete in the first place? Because only
selling new stuff generates revenue! Why declare OS releases EOL? Because
backporting fixes is costly. OK, granted. But then, why does the current
release not run on older hardware? Because IBM wants it to be this way!
Forcing customers to upgrade on many fronts is the way to generate revenue.

And the heap of electronic waste is becoming bigger and bigger…

3 - Service is based on availability of skilled resources. Sorry, we no

longer have anyone in your branch trained on that old model, but thanks for
the maintenance check anyway.

You seriously want to tell me that Power 9 and 10 machines are that much
different so an average guy with a proper left and right hand, and a brain
needs to be specifically trained to be able to replace the main planar for
a Power10 when he's been trained to do this for other, largely same boring
19" hardware?

Power 9 = old model? See subject. We're not talking about someone required
to fix a B30, or so!

4 - New issues are not covered. For example, some hardware in the

industry goes totally berserk after x hundred hours of runtime unless you
put on a certain patch. If they discover something like this, you're on
your own. Such runtime issues are really out there and I'm not talking
about some hidden gotcha feature some people put out there to ensure you're
paying maintenance.

You're talking about one particular issue you encountered with certain
SSDs, do you? How many other examples of such a rare occasion do you have?

And this is an excuse to replace an entire server?

Sorry, can't follow that argument.

:wq! PoC


--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription related
questions.