Boston Linux & Unix (BLU) Home | Calendar | Mail Lists | List Archives | Desktop SIG | Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings
Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU

BLU Discuss list archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Discuss] Server Room Power



On Thu, Oct 13, 2011 at 1:00 PM, Jerry Feldman <gaf at blu.org> wrote:

> One question I have is in planning. One day we received a shipment of about
> 5 or 6 Intel systems from out Toronto office. We ordered a rack, switch, and
> a rack KVM. We initially plugged everything in to a wall outlet. ran fine
> until it tripped a breaker, and the breaker box was not in the computer
> room, but somewhere else on the floor where we had to call building
> management. All I knew that the wall outlets were NEMA 5-20. The solution at
> that time was to take the two 6U monsters and plug them in to separate
> outlets in the ceiling. This worked for quite a while until my boss brought
> in a system he had at home (another 4U Intel whitebox). At that time I had a
> rack power strip, and the power strip popped a breaker, but the wall circuit
> was fine. I then bought another strip to split the load. before all that I
> estimated our power usage by adding up the wattage on the power supplies
> (each was about 700W). Each wall outlet also went to a separate breaker. It
> was at this point when we were getting the HP ESX box and IT somewhat
> dictated that we get 2 240V outlets). Right now I am pulling about 15A (7 on
> one, 8 on the other). But, the critical factor is at takeoff, or when
> starting all the systems, such as after a power fail. You've got all your
> systems spinning up drives and fans. This is what we need to plan. So, I
> would need a rule of thumb that I can take the wattage of each power supply
> and figure out my maximum amps. Had I performed that calculation initially,
> I would have had fewer outages. I can't help when a truck, bus, or tree
> takes out the entire Riverside T station and us :-)
>
>
> On 10/13/2011 12:29 PM, Tom Metro wrote:
>
>> Edward Ned Harvey wrote:
>>
>>> Hold it.  P=VI is a DC rule.  Power is more complex in AC.
>>> What's the difference between VA and W?
>>>
>>> If you have inefficient power supplies, you might be overpaying 30%
>>> for power.
>>>
>> You're referring to power factor:
>> http://en.wikipedia.org/wiki/**Power_factor<http://en.wikipedia.org/wiki/Power_factor>
>>
>>   The power factor of an AC electric power system is defined as the
>>   ratio of the real power flowing to the load over the apparent power in
>>   the circuit,[1][2] and is a dimensionless number between 0 and 1
>>   (frequently expressed as a percentage, e.g. 0.5 pf = 50% pf).
>>   [...]
>>   Circuits containing purely resistive [loads] have a power factor of
>>   1.0. Circuits containing inductive or capacitive elements (electric
>>   motors, solenoid valves, lamp ballasts, and others ) often have a
>>   power factor below 1.0.
>>
>> So when PF=1.0, VA==Watts. The better the quality of your power supply,
>> the closer its PF will be to 1.0. In the last decade it has become
>> common for name brand computer power supplies to specify a PF as a
>> selling point.
>>
>> See also:
>> http://en.wikipedia.org/wiki/**Switching_regulator#Power_**factor<http://en.wikipedia.org/wiki/Switching_regulator#Power_factor>
>>
>> for discussion of PF with respect to computer power supplies.
>>
>>
>>  When you're talking about 208, you're talking 3-phase.
>>>
>> You can attach single phase loads to a multi-phase supply, as long as
>> they are balanced:
>> http://en.wikipedia.org/wiki/**Three-phase_electric_power#**
>> Single-phase_loads<http://en.wikipedia.org/wiki/Three-phase_electric_power#Single-phase_loads>
>>
>>
>>  If you want to use 3-phase 208, you need a special power supply in the
>>> server.  Generally you don't have such a thing...
>>>
>> Old power supplies used to have a 120V/240V mechanical switch. Most
>> modern switching supplies will work fine with any input voltage from
>> like 90V up to 250V (check your supply specifications). The ability to
>> handle a wide input range is a byproduct of the switching regulator
>> design:
>>
>> http://en.wikipedia.org/wiki/**Switching_regulator<http://en.wikipedia.org/wiki/Switching_regulator>
>>
>>  -Tom
>>
>>
>
> --
> Jerry Feldman<gaf at blu.org>
> Boston Linux and Unix
> PGP key id:3BC1EB90
> PGP Key fingerprint: 49E2 C52A FC5A A31F 8D66  C0AF 7CEA 30FC 3BC1 EB90
>
> ______________________________**_________________
> Discuss mailing list
> Discuss at blu.org
> http://lists.blu.org/mailman/**listinfo/discuss<http://lists.blu.org/mailman/listinfo/discuss>
>

NEMA 5-20 is both the connector type, but also tells you it's a 20amp
circuit.  To be safe and up to fire code you CANNOT use more than 80% of
your available power.  So for a 20amp circuit you can use up to 16amps.
 Keep in mind if you have two different 20amp circuits to distribute the
load of a rack of servers and you're hovering at 16amps (actually anything
more than 8-9), when you lose power on circuit 1, you will trip circuit 2
because it cannot handle all the load.  Remember that dual power supply
servers draw half the load from both power supplies, so when one fails it
draws full load from the circuit that's still up.

Also, calculating server amperage is not an exact science.  Even though
Watts = Amps x Volts, it's possible that your server with two 1000 watt
power supplies is only drawing 250 watts (around 2 amps) total right now,
but when it first boots up it will spike up while all the fans and drives
rev at full RPM and can draw around 5-6 amps.  Usually after a few minutes
the amperage goes back down but it can spike back up if there is heavy
usage.

One common mistake I've seen, is when they are installing their servers
(usually one at a time), they power on servers and continue to watch the
amps until they reach their 16amps.  So for the time being their servers are
online.  The problem they have is after a power failure they turn every
server on one immediately after another which causes a huge spike and trips
the circuit.

A long time ago I got fed up with trying to calculate amperage, so I
invested in a clamp on ammeter.  Then I test my servers when I get them and
record the high and average usage.  I also use APC rack rebooters which have
a display of how many amps per circuit (not per server) I'm using and have
SNMP capabilities so I can run Nagios for alerting and Cacti for trending.

Matthew Shields
Owner
BeanTown Host - Web Hosting, Domain Names, Dedicated Servers, Colocation,
Managed Services
www.beantownhost.com
www.sysadminvalley.com
www.jeeprally.com
Like us on Facebook <http://www.facebook.com/beantownhost>
Follow us on Twitter <https://twitter.com/#!/beantownhost>



BLU is a member of BostonUserGroups
BLU is a member of BostonUserGroups
We also thank MIT for the use of their facilities.

Valid HTML 4.01! Valid CSS!



Boston Linux & Unix / webmaster@blu.org