Path: csiph.com!eternal-september.org!feeder.eternal-september.org!nntp.eternal-september.org!.POSTED!not-for-mail From: TheLastSysop Newsgroups: comp.os.linux.misc Subject: Re: Redundancy/Survival Date: Tue, 02 Jun 2026 18:25:37 GMT Organization: The Null Device Restoration Society Lines: 74 Message-ID: <333de5e6978df739348e@dev.null> References: <10v55mv$2co0n$1@dont-email.me> <20260526161738.00004146@gmail.com> <10v677l$2jh1c$3@dont-email.me> <10vb4pn$3tios$1@dont-email.me> <10vitqf$1us3j$1@dont-email.me> <10vkj3a$2dpu1$1@dont-email.me> <2i61fmxegm.ln2@Telcontar.valinor> <10vn6ve$33tsa$5@dont-email.me> Injection-Date: Tue, 02 Jun 2026 18:25:38 +0000 (UTC) Injection-Info: dont-email.me; logging-data="3295225"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1+9VO74iTApHZTpACW1tFCFINn+WN/CGaM="; posting-host="d100ee8f79d7efae5410fccadbdcc1df" Cancel-Lock: sha1:iF1Cl54ilKXbO3PoZy4J7Nl2uwk= sha256:QBkMLbTh34x4901+KrYg3TtBkEsfxz3fs1kXIhRmtzU= sha1:c8z+ROoHKQNHWT7fIfDxnW9owYs= X-Mood: reasonably caffeinated In-Reply-To: <10vn6ve$33tsa$5@dont-email.me> X-Archive-Policy: please preserve the funny parts X-Newsreader: tin can + wet string 0.9.7 X-Operating-System: TempleOS-adjacent abacus cluster Xref: csiph.com comp.os.linux.misc:87384 >On Tue, 2 Jun 2026 18:21:34 -0000 (UTC), Rich wrote: >Carlos E.R. wrote: >> On 2026-06-01 20:30, Rich wrote: >>> Carlos E.R. wrote: >>>> On 2026-06-01 05:20, Rich wrote: >>>>> c186282 wrote: >>>>>> I've EXPERIENCED tower networks Going DOWN ... they have maybe >>>>>> three days worth of power backup. Then it's 1826 again. >>>>>> >>>>>> But copper KEEPS WORKING. Simple and sure. >>>>> >>>>> ... >>>>> >>>>> If you applied the identical regulations to fiber, it too would >>>>> (and could) be just as much a "it KEEPS WORKING" system as the old >>>>> copper POTS system was. All the fiber would need to be all but >>>>> identical is for a pair of power conductors (yes copper wires, >>>>> since glass happens to be an electrical insulator) to be run along >>>>> with each bundle, and for the demarc terminals in each home (plus >>>>> one of the phone handsets) to be powered from the fiber bundle >>>>> power conductors. >>>> >>>> That would not happen. >>> >>> Agreed. I highly doubt such would ever occur to the regulators as >>> well. My point is, it would be /possible/ for the fiber bundle to >>> carry a single copper pair used only to provide the small power >>> necessary to power the end point just like old style analog POTS >>> phones were powered by the line from the switch. >> >> Thing is, they would have to power the ONT or the router, and do so >> for all customers the full time, even if they are not using their >> phones actively at the moment. That is no small power, it can be >> 2..3 amps at 12 volts per client (based on the ratings of the power >> wall wart of my router. > >Don't look at the label on the wart as the consumption of the device. >The wart's labels are what it is rated to output without lighting your >house on fire. If you measure the actual draw of the connected >devices, you'll often find it is much less than the wart's max 2..3 >amps. Often substantially less. > >In any case, that is also fixable. Run the power conductors at a >higher voltage (48v -- the POE voltage) or even whatever your mains >voltage is (220 in Spain?). Then the current in the big cable is much >reduced, even with every end point 'running' and drawing power. It can >be done, but it is also very much a not likely to be done thing as >well. > >Of course this would likely mean fewer "end points" on each arm from >the local switch, as no matter how high one runs the voltage on the >power pair, there is a limit of how many demarc's it can power at once. > >Alternately, each "fiber line" could include a pair of power conductors >to power that one end-point (again likely running a higher voltage to >cover for resistive losses in the long wires). > >But also, just an 'imagined possibility'. No regulator today would >think of trying to impose any such requirement anyway. For a home setup the useful first step is to measure the real load, not size from the wall-wart label. A cheap plug-in wattmeter on the AC side, or an inline DC meter if the router/ONT uses a barrel connector, will usually show a much smaller steady draw than the adapter's maximum rating. If the goal is keeping voice or basic connectivity alive through short power cuts, a small DC UPS or a modest AC UPS sized from that measured load is a lot simpler than trying to make the access network provide line power. Then check whether the ONT, router and any ATA/DECT base all need backup; missing one box in that chain is the usual surprise. -- TheLastSysop "I survived the great rm -rf / rehearsal and all I got was this .signature."