[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [datacenter] peak capacity planning for power vs. cooling
To elaborate further, you normally don't add up every possible piece of
equipment because in practice you are never at 100% capacity (electrically).
So generally there is always headroom. Even with all identical pieces of
equipment there might be a few second differences between achieving peak
load which is all you need to keep the breakers from flipping.
If you are overly concerned about this, you can get breakers that have
power-on delays of a certain number of seconds to avoid this problem -- this
is sometimes less expensive than overengineering your power. YMMV.
In practice, I have never seen a data center of a reasonable size that was
ever at 100% power capacity. I have seen very small ones (<1000 sq feet)
that were at 105%. Neither flipped breakers when the power went out and came
back on. The smaller one (I am thinking of) had UPS systems that took
slightly different amounts of time to power up and heterogenous pieces of
equipment as loads.
DJ
> -----Original Message-----
> From: Martin Hannigan [mailto:[email protected]]
> Sent: Wednesday, January 15, 2003 4:52 PM
> To: Ofer Inbar
> Cc: datacenter
> Subject: Re: [datacenter] peak capacity planning for power vs. cooling
>
>
>
>
> IIRC -- Break it all down to watts, divide by usable square footage.
> Watts-Per-Square-Foot
>
>
>
>
> Regards,
>
> --
> Martin Hannigan [email protected]
> Fugawi Marine Divers http://www.fugawi.net/
>
> On Wed, 15 Jan 2003, Ofer Inbar wrote:
>
> > On Wed, Jan 15, 2003 at 03:40:58PM -0500, Martin Hannigan wrote:
> > > For technical purposes, I'd say plan is based on the
> > > watts-per-square-foot number.
> >
> > Elaborate on that?
> >
> > --
> > -- Cos (Ofer Inbar) -- [email protected] 781-273-2380
> > -- Permabit, Inc. -- [email protected] 617-252-9600
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [email protected]
> > For additional commands, e-mail: [email protected]
> >
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]