RE: "Efficiency" Was: Meltdown and Spectre - Additional PTFs released

Rob,

You know what I'm after. I want a real world example of the performance
impact of the PTF. If the system is real busy, that says one thing. If
it's like yours where it's idling most of the time, and still has that
impact, that says something quite different.

As for my customer set we have not seen that type of degradation in
performance as yet but I'm curious about the rest of the world at this
point.

And yes, I tell my customers that the percent of CPU is also a measure of
the percent of investment being used. 75% - 80% is good, 90% is better,
assuming no latent demand and jobs all finish as expected. We rarely see
those types of numbers in smaller customers since the CPU is so big, but in
larger instances, we see it constantly.


--
Jim Oberholtzer
Agile Technology Architects


-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Rob
Berendt
Sent: Tuesday, January 30, 2018 7:47 AM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Subject: "Efficiency" Was: Meltdown and Spectre - Additional PTFs released

<snip>
If the system was not real busy (meaning over 80% CPU over a 20 minute
observation)
</snip>

What, you don't have management who feels that if you're not at or above
that on a regular basis you're not getting the true efficiency out of the
machine? :-) Luckily not an issue here:
% CPU used . . . . . . . : 11.9

Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1 Group Dekko Dept 1600 Mail
to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com





From: Jim Oberholtzer <midrangel@xxxxxxxxxxxxxxxxx>
To: "'Midrange Systems Technical Discussion'"
<midrange-l@xxxxxxxxxxxx>
Date: 01/30/2018 08:18 AM
Subject: RE: Meltdown and Spectre - Additional PTFs released
Sent by: "MIDRANGE-L" <midrange-l-bounces@xxxxxxxxxxxx>



Tsutomu:

How busy was the CPU and memory prior to the installation of the PTF?

If the system was quite busy before, that tells us one thing. If the
system was not real busy (meaning over 80% CPU over a 20 minute
observation) then that says something entirely different.


--
Jim Oberholtzer
Agile Technology Architects