Re: Giving more than 110%

[Dave Schnee <DSchnee@xxxxxxxxxxxxxxxxxxx>]

Kirk:

What probably happened to you was that you were on a System i5 machine 
with "uncapped processors" and LOTS of unused CPU power in the processor 
pool.  In that instance, you can "borrow" CPU cycles from other 
partitions' allocations if you have your "virtual processors" number set 
higher than the actual processor(s) you are assigned.  Since you must have 
at least 0.10 processor assigned for each virtual processor you assign, 
the most you can get is 1,000% (even if there are more processors still 
idle) because i5/OS will only dispatch as many concurrent jobs as there 
are virtual processors.

In your example, you were assigned 0.50 processing units and could, 
therefore, assign up to 5 virtual processors (simultaneous threads 
executing).  That would allow you to use 1,000% of the CPU you were 
assigned.  If there were not enough idle CPU cycles in the processor pool, 
you would have used less than 1,000% despite your assignment of virtual 
processors.

Had you assigned 1 virtual processor with 0.50 processor units, you could 
have gotten only to 200% CPU usage if there were otherwise-idle CPU cycles 
(due to other partitions not using their full allocations).

There is a Red(something) - either a RedPaper of RedBook - I'm not sure 
which it will become - that is currently being reviewed before publication 
by IBM.  I am one of the authors and there is a rather complete treatise 
on this topic in that writing.  The publication number has been assigned 
as SG24-7174 and the current working title is "Automating Dynamic LPAR on 
iSeries and System i5 utilizing APRM".  I hope it gets "out" soon - this 
is not a well-known concept.

Dave Schnee,
Barsa Consulting Group, LLC

------------------------------

message: 6
date: Sat, 04 Mar 2006 06:43:27 -0800
from: Kirk Goins <kgoins@xxxxxxxxxxx>
subject: Re: Giving more than 110%

When I was in LPAR class we drove a system to 1000% by running multiple 
copies of SETI@HOME in PASE... The system doesn't have to be barrowing 
CPU from anywhere for this to happen. When configuring and LPAR with an 
HMC you tell the system how much CPU, say .5 of a CPU and you can then 
say how many virtual processors you want. These are in .1 or 1/10th of a 
cpu settings. So in my example here you can tell it anywhere from 1 to 5 
virtual processors. If you tell it 1 virtual processor your 
wrkactjob/wrksyssts etc will look normal and when you have used 100% of 
that .5 cpu you gave it then it will show 100% on those screens. If you 
chose to assign 5 virtual processors then it would show 500% when you 
have maxed out the .5 of cpu.

Why would you set more than "1'" virtual processor? Some type of work 
use lots of small chunks of cpu or that have lots threads, these "may" 
benefit from multiple virtual processors...

Mike the Code Poet wrote:

>That's nothing, I believe at one point our 570 was giving us 250%.
>
><WARNING, WARNING... Cheesy humor follows>
>
>... now that is a dedicated box! ;-)
>
>On 3/3/06, Peter Dow (ML) <maillist@xxxxxxxxxxxxxxx> wrote:
> 
>
>>Hi Everyone,
>>
>>I'm sure this has something to do with LPARs, but based on this, one
>>could say that iSeries give more than 110% effort when trying to finish
>>a job -- note the CPU% from this WRKACTJOB:
>>
>>                             Work with Active Jobs
>>DEVELOP
>>                                                             03/03/06
>>14:15:47
>>CPU %:   165.2     Elapsed time:   00:00:07     Active jobs:
>>598
>>
>>
>>Type options, press
>>Enter.
>>   2=Change   3=Hold   4=End   5=Work with   6=Release   7=Display
>>message
>>   8=Work with spooled files   13=Disconnect
>>...
>>
>>
>>Opt  Subsystem/Job  User        Type  CPU %  Function
>>Status
>>        RECURRING    GMHMTHEND   BCH    ++++  PGM-HSCP800B
>>RUN
>>
>>
>>Peter Dow
>>Dow Software Services, Inc.
>>909 793-9050