On Wed, Oct 15, 2008 at 8:38 PM, Will Murnane <[EMAIL PROTECTED]> wrote:
> On Wed, Oct 15, 2008 at 23:51, Scott Laird <[EMAIL PROTECTED]> wrote:
>>>> Most 3.5" drives want
>>>> about 30W at startup; that'd be around 780W with 16 drives.
>>> I'm not sure what kind of math you're using here.
>>
>> See
>> http://www.seagate.com/staticfiles/support/disc/manuals/desktop/Barracuda%207200.11/100452348b.pdf
>>
>> Seagate claims 2.8A @ 12V per drive at startup. That's 33.6W. The
>> operating draw is way lower--the last time I measured my E2160 + 10
>> disk system drew around 130W while idling and not a whole lot more
>> while active.
> 2.8A at startup sounds about right. I still like my 20W-per-drive
> rule of thumb for home systems; well-built power supplies can
> momentarily deliver more than design power to account for spinup, and
> in the event of a failure one can RMA or replace the dead supply with
> little impact. But my original point was that 16 times 30 is 480, not
> 780; 300 extra watts for a system is something I'd expect to see from
> a 16-core machine or something. There's some advantage to allowing
> overhead in one's power supply choice (capacitors aging), but 40% is a
> bit much.
Your rule of thumb is probably good logic - although one could debate
the number forever. Your 20W number seems reasonable to me - and +1
that a premium PSU can tolerate some abuse for the 30 Seconds or so
required during drive spinup.
> For work use, of course, design for redundancy and uptime. But for
> most things I'd use a home server for, spending that kind of money is
> a waste. Money one would spend on a gargantuan power supply to sit in
Another issue here is that the PSU will only be efficient over a
certain, limited, range of power output. If you install a monster PSU
(say 1000Watts) and your average system power consumption is 300
Watts - then the PSU may not be operating in the power range where its
efficiency is optimum. This inefficiency will be reflected in your
power bill.
> your basement is generally better invested in an off-site backup, even
> if all the difference buys is a DVD that you leave at Grandma's house
> when you visit.
I detect some technical confusion reading this thread. A lot of
posters are talking watts (power supply watts), one poster is talking
watts from the wall (plug) and no-one is quoting power consumption for
the CPU - which can vary greatly depending on the CPU model # and what
the CPU is doing at any given time.
The important point is to look at power consumed in *amps* from the
12volt section of the power supply. The CPU(s) are pulling most of
its power from the 12v section of the PSU. So, total up the the
required amperage from the 12v section of the PSU (including
everything that draws power from it[1]) and then carefully examine the
available power from the 12v output(s) of the PSU. Ignore total power
supply wattage.
One potential issue is that some PSU manufacturers don't publish the
power (in amps) available from the 12v section of the power supply -
except on many of the top-end, premium PSUs where there will probably
be 2 separate 12 outputs and the data is generally available - and
then you'll have the challenge of balancing the load across both 12v
sections.
The exception to the "rule" of multiple 12v output sections is PC
Power & Cooling - who claim that there is no technical advantage to
having multiple 12v outputs (and this "feature" is only a marketing
gimmick). But now that they have merged with OCZ - who always claimed
that there are advantages to multiple 12v output sections ... I'm not
sure where they stand today. In any case the PC Power & Cooling PSUs
are premium, reliable, high performance parts in my personal
experience - altough their "Silencer" products are far from silent in
my experience! :)
In the "old" days of (parallel) SCSI, you had staggered drive startup
available (based on the drives SCSI ID). In that respect we seem to
have taken a step backwards with SATA drive technology.
[1] 12v power is used for the CPU, fans, graphics cards, controllers,
disk drives .... etc.
> Will
>
> PS: None of this stops me from wanting the work-grade stuff at home...
+1
Regards,
--
Al Hopper Logical Approach Inc,Plano,TX [EMAIL PROTECTED]
Voice: 972.379.2133 Timezone: US CDT
OpenSolaris Governing Board (OGB) Member - Apr 2005 to Mar 2007
http://www.opensolaris.org/os/community/ogb/ogb_2005-2007/
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
