AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:
CFG:
PT:0
S:andy.sharp@lsi.com
RQ:
SSV:mhbs.lsil.com
NSV:
SSH:
R:<Wayne_Weilnau@Dell.com>
MAID:2
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/LSI/INBOX	0	07E32F241046DA418A0381C1225BFC0119C2C29021@AUSX7MCPS301.AMER.DELL.COM
X-Sylpheed-End-Special-Headers: 1
Date: Mon, 5 Apr 2010 15:50:47 -0700
From: Andrew Sharp <andy.sharp@lsi.com>
To: "Wayne_Weilnau@Dell.com" <Wayne_Weilnau@Dell.com>
Subject: Re: Power measurements
Message-ID: <20100405155047.59b5beba@ripper.onstor.net>
In-Reply-To: <07E32F241046DA418A0381C1225BFC0119C2C29021@AUSX7MCPS301.AMER.DELL.COM>
References: <mailman.5.1270486803.20622.linux-poweredge@dell.com>
	<20100405120736.626d6dba@ripper.onstor.net>
	<07E32F241046DA418A0381C1225BFC0119C2C28E22@AUSX7MCPS301.AMER.DELL.COM>
	<20100405135816.06fe27ff@ripper.onstor.net>
	<07E32F241046DA418A0381C1225BFC0119C2C29021@AUSX7MCPS301.AMER.DELL.COM>
Organization: LSI
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

On Mon, 5 Apr 2010 16:17:50 -0600 "Wayne_Weilnau@Dell.com"
<Wayne_Weilnau@Dell.com> wrote:

> First off, I'm purely a software engineer so .......  Measurement of
> the power draw is done within the power supply.  In the case of a
> blade, there is a sensor that measures the power coming into the
> blade.  This is then converted to wall draw since we have a fairly
> good idea of the efficiency of the power supplies.  I did not
> understand your statement "...seen several cases where they don't
> match".  Are you saying you used a meter to measure power draw and it
> did not match what was reported by OM or iDrac?

Yes, that's what I'm saying, but with one, possibly crucial,
exception:  I don't recall using OM or iDrac.  IIRC, I just used
a sensor driver of some kind.  This was a couple of years ago, so....
What I remember is that I arrived at the conclusion that the sensor
must be on the DC side of the power supply, and therefore just on the
motherboard.  I suppose it's possible to have it on the AC side,
however, I wouldn't have expected that.

But if OM compensates, then that's pretty good, I have to give kudos
for taking that extra step.

Cheers,

a

> Wayne Weilnau
> Systems Management Technologist
> Dell | OpenManage Software Development 
> 
> Please consider the environment before printing this email.
> 
> Confidentiality Notice | This e-mail message, including any
> attachments, is for the sole use of the intended recipient(s) and may
> contain confidential or proprietary information. Any unauthorized
> review, use, disclosure or distribution is prohibited. If you are not
> the intended recipient, immediately contact the sender by reply
> e-mail and destroy all copies of the original message.
> 
> 
> -----Original Message-----
> From: Andrew Sharp [mailto:andy.sharp@lsi.com] 
> Sent: Monday, April 05, 2010 3:58 PM
> To: Weilnau, Wayne
> Cc: linux-poweredge-Lists
> Subject: Re: Power measurements
> 
> On Mon, 5 Apr 2010 14:39:34 -0600 "Wayne_Weilnau@Dell.com"
> <Wayne_Weilnau@Dell.com> wrote:
> 
> > Power is actually reported in wall draw.  This includes drives,
> > power supply inefficiencies, etc.  In the case of a blade system,
> > the power is measured at the blade but is converted to wall draw.
> > It includes all devices on the blade, but does not include any
> > chassis infrastructure such as fans, IO modules, Chassis management
> > controller, etc.  The Chassis management controller also reports
> > power consumption.  This also is reported in the context of wall
> > draw.  Since it monitors the entire chassis, this number includes
> > fans, blades, CMCs, IO modules, power supply inefficiencies, etc.  I
> > will respond to some of the other issues raised by John Lloyd
> > separately when I have time to verify a few items.
> > 
> > Wayne Weilnau
> > Systems Management Technologist
> > Dell | OpenManage Software Development
> 
> Hm.  Well, I've seen several cases where they don't match, and none
> where they do.  (None involving blade systems).  Is it possible that
> some systems have better sensor placement than others?
> Unfortunately, I don't have any specific information because this was
> at a previous place of employ.
> 
> Cheers,
> 
> a
