[180270] in North American Network Operators' Group
Re: stacking pdu
daemon@ATHENA.MIT.EDU (Owen DeLong)
Fri May 29 21:40:39 2015
X-Original-To: nanog@nanog.org
From: Owen DeLong <owen@delong.com>
In-Reply-To: <5568DEA0.6010802@monmotha.net>
Date: Fri, 29 May 2015 18:34:06 -0700
To: Brandon Martin <lists.nanog@monmotha.net>
Cc: nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
> On May 29, 2015, at 2:48 PM, Brandon Martin <lists.nanog@monmotha.net> =
wrote:
>=20
> On 05/29/2015 05:29 PM, William Herrin wrote:
>> On Fri, May 29, 2015 at 4:32 PM, shawn wilson <ag4ve.us@gmail.com> =
wrote:
>>> Is there a way to stack PDUs? like, with 30A 220, we need more plugs
>>> than power but I'd like them to communicate to make sure we don't =
over
>>> power the circuit. Do any APC or Triplite systems support this?
>>=20
>> Isn't it against the NEC and the fire code to stack power strips? We
>> all do it, but isn't it against code?
>=20
> It sounds like he's asking for a PDU specifically intended to be used =
with managed expansion/stacking type modules. That would be permissible =
per NEC and most local fire codes as the device would be specifically =
listed for such usage. I've seen something like this inside =
blade/modular racks like IBM Flex systems, but unfortunately I don't =
know of any such off-the-shelf parts.
>=20
> It would seem entirely reasonable to take a 30A/240V feed on a 6-30 or =
14-30 type connection and break it out to numerous either C13 or =
6-15/5-15 style sockets, possibly on a modular basis using some sort of =
proprietary power+data connections for the breakout. E.g. if you needed =
to mix 120V and 240V devices in a rack, you could get 5-15 and 6-15 =
breakouts with each individual outlet switched and maintain whole-rack =
metering.
>=20
> Sounds like a really handy device, if one exists COTS.
> --=20
> Brandon Martin
If I read it correctly, stacking PDUs is perfectly acceptable so long as =
the breaker feeding each subordinate PDU is sized no larger than the =
maximum current that can be safely delivered to that PDU.
However, adding extension cords or other =E2=80=9Ctemporary power =
extenders=E2=80=9D on a permanent basis is also unacceptable and most =
PDU stacking of the type being contemplated here involves deploying the =
subordinate PDU in a way that ends up being essentially an =E2=80=9Cextens=
ion cord=E2=80=9D for this purpose.
Owen
