This article is great overall, but there is some stuff you should do differently, if you're wanting to follow in the author's footsteps:
1. Just use UPC connectors everywhere unless you have really good reasons not to.
2. Use single mode fiber (SMF) everywhere, including internal wiring. SMF is infinitely scalable for future needs, multimode (MMF) is not. These days the cost differences for cabling and optics is negligible. In fact, SMF cabling is often cheaper than MMF. And no, you generally won’t have to worry about attenuating signal.
Also, in my extensive experience, FS.com optics modules are typically better manufactured and more reliable than OEM modules.
I resisted the "cheap chinese" modules for a long, long time, but then discovered that most of my uplink carriers used them.
It's been years now. I've had many Cisco OEM, Dell OEM, and a couple from another 3rd party manufacturer go bad, but I have yet for one FS.com SFP to go bad. I've got several hundred deployed in a wide variety of environments, from great (cool, clean areas) to bad (hot, dirty areas).
Fiber is, as runjake mentioned, very very cheap. The raw cost is less than copper, and that is part of why I ran so much of it in my build. (I am not the OP). I did however include some multimode in that equation only because some AV equipment uses is, and not all of them use replaceable optics modules.
I would strongly agree that if you are running fiber, run single mode everywhere. It is amazing how far singlemode fiber can go compared to copper if you are willing to spend money on the endpoints.
What kind of equipment outlay was necessary if you don't mind my asking? I.e. the splicer the OP bought seems a bit expensive and impractical to own compared to an ethernet crimper, etc.
I just ordered the length I needed from fs.com. Granted I only have a single run between 2 switches home<->garage. I ran the fiber in a PVC conduit. Put a 4”x4” pvc junction box in the run with a couple feet excess coiled inside because my measurements were a little off.
At the termination of the runs I have fiber keystone jacks and I can extend/replace the last bit if I need to move equipment.
I had a hard time finding UPC wall outlets since most that are available in the Swiss format are for FTTH use, so they all have APC LC connectors on them. The keystone versions are available as UPC but they lack any fiber routing in the back.
My setup is much simpler - I get 25Gbit/s from init7 to Mikrotik CCR2004 https://mikrotik.com/product/ccr2004_1g_12s_2xs and then distribute it to 2x1Gbit access points, and use the other SFP28 port to power my PC (in another room) via the Intel E810 card.
I use the Feller fiber extension cable (20m) and bought some cable puller to pull the fiber via the plastic piping, and then finished it with the Feller outlet. Looks rather clean :) Photo: https://pbs.twimg.com/media/FWhHwD0WAAIb3Dt?format=jpg&name=...
PS: ccr2004_1g_12s_2xs is barely coping with 25Gb/s routing. It can do it, but for multiple TCP sessions, a single one chokes the router at something like 20Gbps (down), 15 Gbps (up).
IIRC ~70CHF for the outlet (which only works with the Feller electrical wall outlets AFAIK) + 10CHF (RJ45 module) + 20 CHF (single-port LC/APC module).
The cable extension is expensive, 150CHF or so for 20m, but you'd probably need it for this outlet because it's specially designed so it can be attached to the LC/APC module manually (no splicing). Something like Swisscom's ClickLC, just a different system.
Which is why I'm so annoyed that my employer (which happens to be the largest employer in my entire state) mandates a vendor that costs 6x as much and whose products fail much more often.
I'm really out of the game these days so don't know the state of different options, but to be fair, a decade ago I had a team member try and cheap out on modules and it caused a fairly decent mess. Lots of failures, but not just outright failures, but problems where the link would mostly work, but through a bunch of fiber taps and bypass modules produced a high error rates when installed.
So I'm not saying it's right, but I can at least understand where a mandate like that might come from, if perhaps outdated compared to the current state of the industry.
The details matter. Depending on how you contract for support, etc. 6x on optics may be the logical choice as compared to being in some situation where the vendor won’t fix stuff until you verify that it’s broken with their OEM parts.
If you’re the largest employer in the state, there may be some personal implications to a CIO telling a major OEM to f-off.
I agree there's a lot of useful information in the article. Two things:
1) I question his choice of OpnSense vs. OpenWRT. I've found OpenWRT to be less demanding on the hypervisor/host, and quite scalable.
2) Not about the article, but about your comment on Chinese modules. I've tried Chinese 10Gb SFP+ modules and I've returned all that I tried. They would overheat, and they would not operate up to the specified data rates. The name brands cost twice as much, but they work.
Good advice. I’ve had the same policy with regard to SMF vs MMF for quite a while now, when I saw each new speed dropping in distance with multimode. You just know SMF will do what you need.
We’ve been using the FS modules too for years, they are very reliable. It does totally make a mockery of the vendors who claim they need to lock out other brand’s modules (from their 100% standardised port) for reliability etc.!
3rd party and maybe 1-2 of each type from the vendor (used to show their support that the 3rd party transceiver isn't the problem) is what we usually buy, basically an open secret. Though I've seen 97% off the price for vendor optics, still more expensive than 3rd party
And given some of the inexpensive single mode fiber spools are two fibers, it is probably easier to just have two fibers instead of trying to do 2 wavelengths.
And to be clear, there will be hacks that scale MMF a bit, but it’s not really “infinitely” scalable due to relatively poor attenuation and light mode. Keep in mind, there isn’t one MMF spec, there’s many: eg. OM1 through OM4
> I resisted the "cheap chinese" modules for a long, long time, but then discovered that most of my uplink carriers used them.
MSP ISP here, we just rock the FS optics combined with an FSbox to code them as required. ;)
Just a warning for the weary: don't put a non-Fiberstore optic in their FS coding box -- you'll lock your account for a week. Ask our field team how they know this.
Is there a best practice for which bidi direction to choose for which device type?
i.e. I assume you don't want to just wing it for every link, so do people typically choose to use all A/B on the switch side, and all B/A on the desktop/server side? If I were walking into a business, not having setup their network, what direction would I expect to need on the desktop/server side?
Aren’t BiDi transceivers usually a good bit more expensive, especially compared to how low the material cost of fiber is? What’s your use case for BiDi? I don’t have a lot of first hand experience with this stuff, though. I can easily imagine BiDi being useful to retrofit/repurpose some existing fiber, since labor costs can be a lot.
I can see it being useful for running ultra thin bend insensitive fibre around base boards; just like the FTTR thing Huawei is trying to sell.
I wish the concept was more clearly legal in Australia. I’m pretty sure once you glue the fiber in place it becomes ‘permanent’ and thus has to be done by a licensed cabler with a fiber endorsement.
I can't speak for anyone else, but we wing it for every link. When we light up a new link, we grab a matched pair, randomly select one to put in the switch, and then put the other one downstream.
I always end up green with envy that people seem to have pre-existing conduits in their walls for running new cables. There's no way, in my old house, of doing this without some invasive smashing in to walls.
I'm currently in the process of literally tearing down walls in my old house (built in the 40's). I used to fear smashing into walls. Once you do a bit of drywall and mud yourself, you realize it isn't that big of a deal. Everything I've learned has been off YT videos and there are some excellent channels with all of the professional tips and tricks. It isn't rocket science by a long shot.
I’ve done a lot of drywall work and I still avoid it whenever possible.
Doing basic drywall and paint isn’t rocket science, but getting a professional level fit and finish actually takes a lot of practice and experience. It’s also a completely different experience when you’re trying to match old specialty paints from a previous homeowner, blending repairs to match the wall texture, and other complexities that don’t show up in those basic YT videos. If you’re in a situation with untextured walls, known paint colors, simple repairs, and a forgiving eye, then it’s really not bad, but the complexities can add up quickly.
Even in the best cases, cutting into drywall all across the house to pull wire would be a project I’d try to avoid at all costs in an occupied home. The fast path still requires multiple days to go through all of the drying and re-application cycles to build up all the right layers, not to mention dust control to keep that fine dust out of everything you own.
I just redid my whole kitchen to the studs. You're making it out to be a bigger deal than it is. Now I'm onto a bedroom where I removed two whole walls that had been weirdly built to cut the room into a smaller room. This is going to require a bunch of repair work, all the way to the ceilings.
Certainly, if you're looking for perfectly smooth walls, then you're never going to be happy regardless. My place is Santa Fe, which is perfect... it is supposed to be messy. But this is a place built in the 40's... nothing is perfect. Doing something perfectly would stand out even more.
If you want old paints to match, then just paint the whole wall. That's easier and probably the right solution anyway.
Sure, it’s easy when you have the wall texture that is designed to look like a bad drywall job. If you’re trying to match the wall type found in 90% of American homes, there’s a huge difference in efficiency and quality between a DIYer and a professional mudder.
A good feathering job saves a ton of wasted sanding effort and drywall dust. When I ask GCs and tradespeople which job they would would never do themselves, drywall is far and away the most common answer. It’s the only job I regret doing myself.
From what I've learned on YT, using quick drying mud and mixing it yourself (not using the premixed stuff in plastic containers), is the key to smooth finishes. Mostly because it hardens quickly and doesn't shrink. I used the 40m quick dry and it was quite easy to work with. You're right, feathering correctly... no sanding needed.
Picking the correct timing for the cure time with the job size is almost an art in itself. It’s unlikely one can figure it out without a lot of experience doing it and most folks that don’t do it for a living won’t do it often enough to acquire that experience. It’s a lot of hot mud waste and back and forth trips to the hardware store that’s exhausting for busy folks that have another full time job.
I used a drill with a mixer from a hand blender plugged into it. Made it really easy to make a batch that filled the stainless steel mud box I am using. Even being totally untrained, getting that onto the wall in under 40 minutes was easy. Took a bit more mixing and batches than a pro would take, but all in all... it wasn't nearly as hard as you're making it out to be.
I'm teaching myself that one too. Will replace with luxury vinyl. I just pulled up the oak floor in the bedroom. It had asphalt paper under it. Sigh. Date on the wood was 2001, so not terribly old, but definitely could use an update.
Since the house is old, it is plywood floors and they aren't even enough for vinyl. Going to have to learn how to fix that too. From what I can tell, it is basically just mud like on the walls... but portland cement on the floor.
By removing and replacing entire walls, you are missing out on most of the difficult parts of texture and paint matching (and "just paint the whole wall" quickly turns into "paint the whole room" in some cases). I suppose you could make the argument that we should then just rip out the entire wall to run a single cable, but I personally think that's extreme both in cost and disruption.
I'm glad you had such an easy time of it, but cutting, patching, and matching is a pain in most homes I've done it in. And don't even get me started on lath and plaster.
I didn't say it was easy... it just isn't so overblown "hire a professional" as people make it out to be. I was terrified before I started, but once I watched a bunch of YT videos and then got to work... it ended up being a lot less stressful than I imagined.
Pretty much impossible to do without major strain on a relationship if you live with someone else, and god help you if you have kids. FYI.
Doing it down to the studs and all at once is definitely the easier way to do it. Small patches and piecework suck, but if you’re living in it, especially with others, you don’t always get a say in the matter.
The trick, in my experience, is to use very fast drying joint compound always in powder. You can mix joint compound that will dry faster than you can put it up, which enables you to put multiple coats on. Sanding stinks, but if you've a good hand you can minimize the amount of post skim finish work.
Used this technique to rip out and then match a ceiling in a bathroom, and also to patch up work after some HVAC dudes blew part of a wall out for mini splits.
It was tough to learn as there is a knack, but paid off. Having good hand eye coordination pays off in many ways.
The most important part was learning from an experienced person: "No wall on earth is perfectly flat, you just want it to look flat." Good advice, especially in a town (Philly) where there are old houses with not a square corner in the whole place. :)
Yep, this was one of my great enlightenments: no matter how good or professional a wall looks, it's really just drywall mounted on wood, some cracks sealed, and then painted over.
When the walls are off in the house is the best time to get anything done. Sometimes I don't even want to put the walls back on- for example, an unfinished garage with exposed studs is very convenient.
How much of that is just a cost thing? Is hiring people to do home maintenance inexpensive and easily available where you are? That's usually not the case in American cities.
Some of it is cost. And almost everyone knows a handyman or somebody in the neighbourhood who is an electrician, concrete guy, tile guy, &c. But part of it is cultural. It's considered a form of virtue signalling to have a bunch of servants or workers on your property. Most middle class people also have maids, nannies, drivers, and such.
How much of this is just the hacker news demographic skewing perceptions? For a lot of people here, it makes far more sense to hire someone to do home maintenance while you continue to work on a well-paid technology job. Doing a lot of home repair is a hobby/choice, not something you're doing out of necessity.
For techies, for sure. Also most techies (in my experience) don’t have a lot of exposure or hands on with blue collar type stuff, which most of this is.
Every town in America has at least one, if not several Home Depot’s, Lowe’s, OSH’s, etc. and at least 75-90% of the traffic at any given time is private home owners, not contractors. Americans do a LOT of DIY.
There isn’t any real comparison anywhere else I’ve been able to find, and I’ve looked. Not as much in France, but Germany, Switzerland, Italy, India, Bulgaria (closest to having something like that), Singapore, Malaysia, Japan, etc.
I may someday use shiplap instead of drywall in specific rooms, simply because you can make it so you can take it off and put it back on relatively easily.
One of the big benefits to drywall, that often gets missed in the typical libel it gets, is its fire resistance. It's basically a water-containing gypsum powder.
Some areas where plywood would be preferrable for durability, like a garage, are often prohibited from using something other than sheetrock due to the fire risk. Probably not a big deal unless you have entire walls of it.
I had a (small) fire in the wall of a house in about 2000. Turned out the rubber insulation had fallen off the wiring, in various places there were just pairs of bare copper wires. An electrician friend helped me rewire, and we layed in cat5 and conduit (between storeys) at the same time. Rewiring a house from the 1840s required pulling up floorboards, but no significant work in the walls.
I'm not sure whether older houses are more maintainable by default, or just once you have rocked over some threshold of continuous occupancy, there there are sufficient layers of repair and rework that you just stop worrying about (for example) sawing through a floorboard. While I didn't have to mess with the walls, if I had, I'd just have torn down the existing plaster and redone it, like you say it's not that big of a deal
I agree it isn't rocket science, but plastering if one of those things best left to professionals, much more so than plumbing or electrical for instance. They usually make near minimum wage, are just so much faster and produce a better result.
I've started plastering a few of my walls and after wasting a couple days doing a few square meters ended up hiring a couple guys to do the rest of the house. There is just no comparison in value.
It just take practice, my first try was not perfect but acceptable, the second time was much better and the third on a larger wall with lots of defects looks good, but I probably spent too much time on it, although I am confident that I can now do it much faster.
Another argument to always hire someone else: if you learn to get a nice finish with drywall, you'll start noticing defects in existing drywall. Everywhere.
Highly recommend Vancouver Carpenter on youtube, if you do want to go down that rabbit hole.
That is my experience as well, second or third time I'm vastly improved (not as good or as fast as a pro still, but upgraded to 'acceptable') since I've learned from the largest mistakes. Too bad I have to live with those mistakes though since I did them to my house!
To me it's like painting. Yes, I can paint my walls. But hiring it out is substantially faster and you get people who do nothing but paint all day every day to knock it out.
I used to hang drywall when I was young and got pretty good at making a wall look completely smooth with no seams visible. But it's something that is so much more efficient to leave to the pros because they just do it so much faster than the DIY homeowner can, and it always looks better.
> Also, awkward to talk about how you pay someone a barely living wage as an excuse to not do it yourself.
Guess how much they make if you don’t pay them?
It’s better for everyone if someone who can make more money per hour uses that higher earning power to offload lower income tasks to people in those jobs.
Basic logic fail. I didn’t say it was immoral to do it yourself. I said it was good to support the worker (from both the worker and the payers’ perspectives), not bad.
The view I replied to implied it would be gross to pay someone to make your meal because it’s cheap.
It's really simple: if you want to do something yourself, do it. If you want to pay someone to do it for you, do that. There's many reasons you might choose one route or the other, and they're all valid. Ignore sanctimonious assholes who tell you that you should support workers just because.
We had our loft converted last year, and after watching the builders construct the partitioning walls, I think I would probably manage that. We also had to have some walls removed and re-built and made the house too dusty for the kids to be around. This house is an old English terraced house, and lots of the non-retaining walls are wooden lath and plaster, with horse hair mixed in for good measure!
While it's not rocket science, I always find it to be super annoying to do. Getting things smooth/flat when you're done (generally, I think I wind up doing 4-5 layers of mud) takes a fair amount of time. Plus sanding it down without getting the powder everywhere is difficult; I use a wet sander / sand "sponge", but even then I wind up with some of it spreading around.
Drywall is one of those things I've found it is highly worth hiring out. Even if you hang the drywall, hiring someone to tape and mud and sand can be entirely worth it - they do it day in, day out, can use five minute mud, and can get done in a few hours what would take you days or even a week.
Doing small repairs is easy enough that everyone should at least know the concepts involved.
It takes a lot of practice to be smooth enough at doing it to have a nice result over any large area before the concrete sets. It’s not the type of thing that benefits from going ‘hmm, maybe I should do it this way’ after you’ve started.
Also, it’s much, much easier to do as a small crew - easily 3-5x more productivity per person. Same as drywall. It’s not easy to do as a solo worker, even if you know the tricks.
I’m guessing what people are saying is that they have seen plenty of home jobs where the surface is left with an obvious bulge, hollow, or an hideous line. Some people are very sensitive to surface unevenness, and really notice slight surface imperfections, and some people don’t notice gross errors.
In my very limited experience, many textures are often harder to make look right than flat surfaces. I think I am sensitive to texture variation, because with textures I notice mistakes and I often notice where alterations have been made.
Either way, your Santa Fe look is irrelevant to the topic of fixing problems in walls when they are not textured like yours.
Unless your walls are plaster on lathe. Hah we took 150 large bags of rubble out of the apartment last week. No, it wasn’t a gut reno, just 2 walls and 2 closets.
> I always end up green with envy that people seem to have pre-existing conduits in their walls for running new cables. There's no way, in my old house, of doing this without some invasive smashing in to walls.
I don't own a house and have never owned a house. This will probably sound stupid and I am sorry if this is something obvious.
Is it that my spouse will leave me or my parents will disown me if I run a conduit just hanging on a wall? What is this obsession with hiding all wiring within the drywall? What am I missing here?
Not that I have a leg to stand on because I still can't afford to buy a home outright (and at current rate, never will). I live in an apartment and I don't make any changes to it. I cannot even imagine doing something simple like drill a hole on the door for a doorbell. So, I am definitely hypocritical when I say this. Maybe I am just being salty as a non-home owner. I feel like all of this comes from treating our homes as some kind of liquid asset that we must keep in pristine condition at all times so we can stage it and sell it at a moment's notice.
If you own your own home, why not live in your home like you own your home? Run that conduit across all the walls (and through inside door frames or something like that if you must). If not, do you really own your own home? Why not just live with housing insecurity like I do?
It sounds like you just have different priorities or a different aesthetic sense than folks like me who go to great lengths to hide cables.
To me it just feels good to wake up in a visually simple environment with things out of sight. It feels like magic, in a good way, to be surrounded by performant, reliable and useful technology but to be able to see almost none of it. And I have an automatic negative reaction to visual clutter or conspicuous machinery in my house.
I don't see anything wrong with your way of thinking. I can't easily change how I feel about it and I see no reason why either of us should have to.
Edit: this makes me think about sci-fi spaceships. Battlestar Galactica vs. the Heart of Gold from Hitchhiker's Guide. One has its conduits and controls run every which way and the other is minimalist to a fault. I like them both but would definitely prefer to wake up every morning in the latter!
To be fair, it looks terrible to you (and to me), but there are plenty of people who have no issue with it. Just like there's people who get agida when there's things like toasters that sit out on the kitchen counter, but a lot of people think it's fine.
Just like there are people who don't care about how good or bad code they write as long as it works. Doesn't mean we should accept or promote this as 'good'.
Also, I would say writing bad code makes the code very inaccessible...just like burying important infrastructure in a wall.
Did that water leak last year cause mild growth in the wall? Don't know without some demolition! Want to upgrade your piping or electrical? Gotta destroy some wall, hooray, and then do more work to hide it again. Lunacy.
I disagree, it's much easier to inspect cables or machinery when they're not hidden.
One of the apartments I looked at recently had this really annoying buzz in the kitchen, turned out it was a transformer that the previous owner plastered in. Definitely a big minus for me.
In my current apartment all cabling and piping is outside the walls because it was built before electricity was everywhere and I find it really handy <3. If the look bothers you it's not for you obviously but I don't think it's inherently worse.
I've seen many places that were designed "industrial" with various metal parts (conduit, vents, etc) showing; and the people that owned them were very happy with the look.
Good code also doesn't need to "look pretty", it needs to be correct, and maintainable. Preferably maintainable by a less skilled developer with less domain knowledge. Following that logic, the code that follows the pattern of "highly visible and obvious", like the conduit, is probably the better code.
All that being said, I prefer simple walls, with conduit hidden inside it. I'm just happy to admit that what I like it not universal, and that other have their own styles.
It looks terrible to the vast majority of the population in most developed countries and it will severely impact your home's resale price if you have spiral conduit all over the walls. I can't remember the last time I saw spiral conduit in a home anywhere outside of the garage or basement...maybe a utility closet.
That's a pretty good idea. I'd consider this more if I wasn't just putting some keystone wall plates through the floor into my basement. I'd also have to put trim around a door frame.
3/8” flex isn’t surface raceway, it’s used for connecting a rigid (EMT/IMC/RMC) conduit system to a vibrating piece of equipment like a motor or transformer, among other things.
There are probably residential raceways that blend in better, but I’m not familiar with residential construction.
There are plenty of ways to get a cable from point A to point B inside of a wall, particularly if you have a single story home with an unfinished attic and basement. A spade drill bit, a fish tape, and a multitool/rotozip can get a cable pretty much anywhere if you can drill a hole into the wall cavity from above or below.
A friend of mine lives in a building with concrete floors above and below, all his wiring runs through plastic, surface-mounted, trunking and it is aesthetically disgusting.
I’m an avid DIYer and have no problem buying others’ high-standards DIY work, inspection/permits or no.
If I walked into a listing and saw surface raceway everywhere, I’d only bid what I was comfortable paying leaving room for a gut rehab. It’s evidence of a high level of DGAF at a minimum and likely isn’t the only place that corners were cut.
Do I understand correctly you're proposing to just run it on the outside of walls/ceilings? How would you ever again close doors that it needs to go through?
My aesthetic requirements are basically zero (to the dismay of my partner indeed; we meet in the middle) but I do want to be able to practically use the apartment still.
The interior doors in most houses that I have seen have wooden frames in which it is very easy to make holes for passing cables. Some newer houses might have door frames made of plastic or metal, instead of wood, but even in such cases it is much easier to make a clean hole through the door frame, instead of through a wall.
I have made many such holes in door frames for passing Ethernet cables or TV coaxial cables.
When you do not want to touch the walls, such cables can be routed on the edges between walls and floor and they can be masked by a cover having the wall color, to be invisible, except where passing through a door frame.
Why would you prefer to make a hole in your doorframe rather than the wall? Sheetrock is so much easier to deal with. But perhaps you are referring to a concrete construction, which is uncommon in the US.
Making a clean hole through a door frame takes a few minutes at most, with an electric drill machine, after which you pass the cable through the hole, you crimp the connector on the end of the cable and you are done.
At least in Europe where I live, I have not seen any walls through which you can make clean holes with high probability. Almost always there is some small damage, which must be repaired, e.g. by filling with plaster and/or repainting. In most cases you also need a percussion drill machine, not just a simple drill machine. Here most walls are made either of concrete or of bricks.
So at least in the kinds of walls that I have seen, passing a cable through a wall requires more time, more diverse skills and a larger set of tools and materials than passing a cable through a door frame, which requires only that you have and know how to handle a drill machine and a cable crimping tool.
While common outside the US, most of it in newer constructions tends to be light concrete which is trivial to drill. But even 40MPa concrete is fine with an SDS, rent one if you don't wish to spend money on. In short drilling through outer walls is not difficult either.
> How would you ever again close doors that it needs to go through?
Most interior doors have a gap between the bottom of the door and the top of the floor, and you can cable in the gap if you don't mind the look. If the threshold is carpetted, you can often squish the wire into the seam where the carpet meets the door frame.
You can get one, maybe two, runs into a room that way, so don't put your central switch in a room like that, and if you need more than one drop in a room, put a switch in the room.
Even without carpet, you'd tack it so it stays pretty close to the side of the threshold. 'Nobody' uses the outside inch or so of the threshold, so it's not going to be a tripping hazard. If you are going across the threshold that's different, but in that case, it's better to go over the top of the door frame.
Do it like the real high class cable installers do: run the wires on the outside of the house and just drill through the walls where you want the drops. Looks fine from the inside!
> I feel like all of this comes from treating our homes as some kind of liquid asset that we must keep in pristine condition at all times so we can stage it and sell it at a moment's notice.
This seems like you are missing the point. The REASON that having a conduit hanging on the wall lowers the home value is because it looks bad and ruins the aesthetics… for both the current home owner and a future home owner. The reason the person buying the house would pay less is the same reason me, the current home owner, doesn’t want it… it looks bad.
I understand that some people don’t care about aesthetics at all (I am probably closer to that end of the spectrum than most), but you also have to realize that people care about how their home looks for more reasons than just resale value. Do you think the whole world only makes their houses look nice in case they want to sell?
I wouldn't run conduit on my walls and around doors because I would think it looks awful. People's aesthetics are different. If yours (and those of anyone else who might live with you) allow for such a thing, then sure, go for it.
> I feel like all of this comes from treating our homes as some kind of liquid asset that we must keep in pristine condition at all times so we can stage it and sell it at a moment's notice.
Well, you aren’t wrong. This is exactly the thinking behind that.
The other response is correct that it’s nicer to have it all covered up though. I guess it comes down to a bit of perfectionism. If I stick everything on the walls I will always feel like there is work left to do.
That said, our living room is still (3 years in) a mess of (nice, not industrial) on-wall conduits.
I think it is wrong. Most people want to keep their house looking nice because they enjoy living in a nice looking house… this is the same reason why having a nice looking house sells for more. It isn’t like the whole world somehow decided to value nice looking houses for no reason and so everyone started keeping their houses nice even though no one actually cares.
Imagine going up or down a floor. Really hard to do on top of existing walls, pretty easy with conduits in walls.
For this house, the easiest way to pull wires from one end of the house to another (after the original construction) is to first go down into the crawl space, cross horizontally there, then go up again.
We do have some conduit in place to run cables. One running up to a wall-mounted TV, and another running Ethernet up in to the loft. I don't mind the one under the TV, but I wouldn't like to use it for everything.
Don’t mean to be sexist but in my personal experience wives would consider cable conduit across a nicely painted wall in a nicely decorated room a hard no.
One interesting technique I have seen with fiber is running it along/behind baseboards in old houses. Since it is very small (and you can get even smaller single mode fiber) it can in some cases run along side. Some people remove the baseboard and put it behind it. It creates some routing issues, but might also solve some.
Yea, that can be an issue, although for shorter runs it is surprising how much loss you can have and still work fantastic. There are some very cool bend insensitive multimode fibers that would work well in that case.
I saw a video where someone was doing this and they used little plastic bend guides that were painted trim color.. not perfect but also not as noticeable as I would have thought.
This is one of the things that my father did when he did the wiring for the house ~50 years ago - conduit everywhere. It means things like drilling into the wall is safer (you hit metal rather than wire) and when you want to, you just pull more wire through it.
There are so many parts of it that were "over engineered" for a house 50 years ago that are "oh, that's convenient" now.
Agreed. In most US homes, by code, there needs to be a fire-stop between floors but also for practical reasons there are often cross-bracing mid way up each wall, therefore going from one place to another often involves multiple holes in drywall (which is relatively inexpensive, but also a dusty/annoying nightmare to patch).
If I built a home I'd add conduit, but that's an extremely niche idea.
A flexible drill bit[1] can be used to pass a horizontal member within a wall. You can enter the wall through an outlet hole or use the length of the bit to drill straight up through bottom of the wall, if you’re able to access it from a basement. They usually have a hole in the tip to pull twine along with the head of the bit.
If you have high baseboards/ceiling trim, you can have drywall that ends way above the sill plates which makes it incredibly easy to route cable with the baseboards off.
Great point. With the baseboards off, you can even look inside the wall (e.g. with a phone camera) for clear stud bays and power cables, potentially finding easier runs.
I've watched the video and I still don't understand how you get the wire through the conduit though. Part of the video seemed to happen through telekinesis: the part where a wire magically came out of the wall, allowing you to pull from the other side.
You have to have a wire installed in the first place(i.e coax or just a simple thin but strong wire). That wire is installed when you install the walls/gypsum. Then you use it to pull whatever you need(hdmi etc). When you pull the new wire(hdmi) you attach an additional wire at the tip so that you can pull back your initial wire once your new HDMI is in place.
If you don't have an initial wire installed you can use a magnet kit with a specific wire to pull it through the conduit.(e.g youtube "magnepull"). A camera small comes in handy as well.
If you need to pull wires through the ceiling and you already have recessed lights you can use their wires to pull your new wire.
> You have to have a wire installed in the first place(i.e coax or just a simple thin but strong wire). That wire is installed when you install the walls/gypsum.
Unfortunately, in my experience of 6 or 7 new builds in the last 15 years or so, coaxial cables are usually staples to the studs and useless for this purpose.
>> Unfortunately, in my experience of 6 or 7 new builds in the last 15 years or so, coaxial cables are usually staples to the studs and useless for this purpose.
Well, that's part of the fun/job. Usually you end-up with a few holes as well so you try to get hold of anything that helps you, "one inch" at a time.
As always, the easy way to deal with this is to let someone else do the job(hire a pro/custom installer).
I'm not sure which video you mean, but you use either a fish tape or a pull line.
I know that Verizon also has fiber with a stiffer jacket that they can push through conduit a pretty good distance. I know this because I helped the Verizon installer run FIOS at my daughter's apartment in a building that was over 100 years old. The conduit had a crimp in it so it took the two of us a while to get past that.
Probably referring to this demonstration video regarding the FTTH Squeeze plate where the pull line comes flying through the conduit: https://www.youtube.com/watch?v=ARSpp4B9-X4
You can attach a string to a foam ball/plunger type thing and then suck it through with a vacuum from the other side. Then, fix your wire to the string and pull it back in the other direction.
A plastic bag tied to pulling twine works really well and is less likely to get stuck than a ball or plunger. I always leave a length of twine in my conduits so I only have to use the vacuum method once.
Indeed the plastic bag trick works very very well in a lot of different use cases. I was able to pull twine over 500 feet with this technique while doing my driveway conduits.
I had about 50 feet and 270° of bends in some PVC conduits I'd buried under a section of yard. I'd gotten everything done and asked my wife to come help me get the poly string and then mule tape into the conduits. The poly line practically leapt into the conduit chasing the plastic bag. It was one of the greatest ratios between how hard I thought something might be and how hard it actually was.
Pulling a bit of twine through from my pool over to where I had installed my equipment pad went much the same way, and I was too dumb to think about using a plastic bag ;). Just stuffed a bit of twine in one and and left it loose, walked over to the other and and put a shop vac on it, and -whoosh- here came the string. Felt like cheating. It was only a small conduit, though, 3/4 inch, so it had pretty good airflow.
I had my house wired up with 30 Cat 6 drops (3 being in-ceiling for AP's) and in the end it was simpler to go up to the loft (2 story house) and come down that way, with all of the cables then being bundled up and coming down through a built-in-wardrobe into the garage below.
It required a bit of patching here and there where a small chase had to be chisseled out to get around a cable going through a patch of plaster used for dot and dab (technique to quickly plasterboard/drywall onto brick), but with conduit it would have been a much simpler job.
With that drill bit you won't be able to get far with concrete brick walls...
I hired an electrician to drill the holes for me earlier this year and it took him over 10 minutes per hole to get through the 60cm (2ft) reinforced concrete slab for running cables between two floors.
Honestly, 10 minutes to go through 2 feet of concrete is damn impressive. I had to put a couple 50A electric circuits through my foundation a few months back and I'm pretty sure we spent more like 45 minutes getting through. And we didn't even hit any rebar, which is a minor miracle itself.
I just spent a couple of weeks putting conduits in my brick walls. I couldn't reuse conduits in place for power (regulations, EU) and they are too small anyway. I used Cat6A, not fiber. I don't think an ISP will ever bring fiber to my home. Cat6A is 10 Gb/s so it should be OK for a NAS for a long time. I only have 1 Gb/s network cards now.
I'm moving into my new home (which was built in 1962) in the next month or two. My plan is to run cat 6 throughout, before I get the bulk of my stuff/furniture in. Hopefully I won't need to smash into too many walls but there will be a few. Definitely worth it in the long run though.
There are so many things I think about "if I had just done X in those few weeks before moving in the furniture".
This is an opportunity - you might want to brainstorm X.
anything in the kitchen (before you start depending on it for food and/or fill the fridge with perishable stuff)
mindfully adding specific storage. Could be the difference between moving in and piling all your stuff in random places.
related: garage - before it becomes the "catch all" for unresolved boxes of stuff you move with. A friend said overpaying closetworld for cabinets around the perimeter of his garage was money well spent.
Sorry only just saw this. Thanks for the tips. Thankfully I'm in no rush to move out of my current home (I'll have a 2 month overlap) so plenty of time to get the 'infrastructure' stuff done first.
I wish my house had conduit. To date I've never seen (much less lived in) a house with condiuit.
I live in a split level. The company I hired flat out refused to do a cat6 drop to my upstairs office, after spending nearly 40 minutes getting wire up a single story.
I'm guessing you're in the UK. Even in my relatively new house (80's), almost all cables are chased in, no conduit anywhere. Standard construction in the UK could be improved by just doing basic obvious things that other countries do.
Likely the US. Unless it’s a very new high-end house or you got lucky and bought the house a contractor built for themselves to a high standard, they did the cheapest thing the electrical code allows them to get away with, which certainly wasn’t conduit in the 80’s, unless that was local code somewhere. Certainly not in CA or FL.
Renting a place built a few years ago in NC. It has only one conduit, it runs from the networking panel to the outside where fiber comes in. Unfortunately there was nothing left in the empty conduit to pull things through with and the installer didn't want to spend half a day with a fishtape, so it didn't even get used. Companies really couldn't care less about doing things right w/ regards to people's homes unless there is some direct monetary consequence for them.
No conduit in NC in my 2004 built house, with the exception of specific chase pipes I asked them to install from the basement to the attic. The drywall was already up to get them to do much more than that. I'm guessing all the cat5 is stapled into place behind the drywall. Fortunately between the attic and the ceiling-tile basement I can get to where I need, it's just a PITA because of fireblocks. So many fireblocks.
I did an extensive amount of both fiber and CAT6A in my recent house build, totally about 21 miles together. I did a mixture of single mode and multimode only because of some AV device support for multimode, and because all of it was so cheap compared to the labor and time.
I ended up with significant fiber runs between the MDF and IDF closets, the server room, and all of the AV endpoints which is a typical layout. I did run single mode fiber to all 18 wireless AP access-point locations, but I suspect those will not get used much given I also have 2 CAT6A shielded wires to each AP that can support 10G AND provided power. I do use a direct fiber connection in my office to provide 10G to my network core from that desktop, but that could have been done over copper as well.
Perhaps the biggest gain over copper is my switch interconnects which are 40G from the upstairs closet to the server room. If needed it would be easy to upgrade those to 100G, and since there are 24 strands available you really could expand as far as you are every going to need to go with WDM and the like.
The one really good use case for only fiber was running it down my driveway to our gate, which is about a 1/4 mile in distance. That is something that would be more challenging to do with copper. [although I do have copper installed].
From my perspective the cost of running fiber was almost 0 compared to the entire project because so much more cost is involved in path finding and clearing ( cutting holes, etc), and the actual pulling. I was very fortunate to have a large group of friends who spent a couple of days working with me to do the large pulls, combined with many weeks of evenings and weekends doing the rest myself. Many hands makes that task much much easier.
Not surprised to see you on here. I was the guy who got into the discussion about the merits of single mode vs multi mode with a few of your friends on Facebook a while ago. I was in the multi mode camp mainly due to cost reasons whereas they were in the single mode camp for future proofing reasons.
That said, the cost gap between MM and SM optics today is much smaller especially for 10G (and for the fiber itself it seems to have even reversed), so single mode definitely makes more sense. For some reason, MM fiber is still widely deployed in certain niche use-cases though. Not quite sure why.
For example, I know that it's used on modern military jets for their 10GBASE-SR networks. I wonder if it has something to do with being able to repair terminations in the field? I know MM is pretty forgiving that way. Or maybe it's just another case of them adopting whatever was popular at that exact moment.
It was surprising the amount of AV related gear that was multimode specific! The cost differential has dropped so much now you can run either at low cost so if you have the space you can run both.
MM is easier to do mechanical splices on, and does seem to be overall less sensitive to damage. If your need is 10G and less than 300m it works well.
As an aside I ran a USB over fiber repeater that was multimode specific by accident I used a single mode fiber (same LC connector), and it worked perfectly.
It was a new house. The 18 AP locations serve a couple of different long run purposes and reasons:
The first is to have direct nearly-line of sight locations for the majority of the house which is primarily to support higher frequency(60GHz) and beyond which have poor wall penetration.
The second is to provide good wifi coverage due to signal attenuation from the construction techniques. I did many of the walls in double 5/8s drywall, some in green wall, all insulated, and all solid hard doors. Steel cross structures as well. As a result wifi propagation is surprisingly bad across rooms.
The third is to provide lots of locations to aid in flexibility of having the best locations. I do not use all 18 locations right now, but I may use more in the future.
It is also a somewhat large (>10,000 sq ft, ~ 1000 sq meters) house, so that facilitates a need for a bit more coverage.
I'm using Ubiquity Unifi 6 Pros right now, but am going to switch to the Enterprise ones (with 6Ghz) at some point soon. They are all integrated to the same controller (a UDM Pro), so all SSIDs I use are available everywhere. I even have a couple of them down the driveway all the way to the gate so I have wifi coverage all the way out.
> I did many of the walls in double 5/8s drywall, some in green wall, all insulated, and all solid hard doors
I gather you did this for noise isolation? And by green wall do you mean green glue between the drywall layers? I'm only familiar with this for home theater construction, not an entire home.
For those wondering, as the OP didn't mention it super-clearly imho - but his internet provider is able to do 25gbit [1] for CHF 69 a month (+ CHF ~340 set up cost)
CHF 1 ~= USD 1 these days
Question for OP:
Why did you build such an expensive "router" - you justify it by saying you also want to run PiHole, but you probably could have done this whole setup much cheaply
The router is more of a server. Pihole is just an example (and yes a $35 pi would be plenty for it) but I have many old servers running all kinds of other things so I decided to update and consolidate some of this.
Has the 5800x and opnsense been able to successfully route 25Gbit/s? It's very rare to find anyone running even a 10Gbit link on opnsense/pfsense so it's difficult to know how it performs on high speed links with modern hardware. Would love any details you could share on this! (Might be worth updating your blog post with them)
I made this telegram bot[1] a while back that uses rpilocator.com and will notify you as soon as one is available from an official reseller. Those resellers should be selling them at the regular price. But yes, if you want to buy one right now, it won't be $35.
Meh. It’s (supposedly) a transient situation that’ll resolve in due time.
More importantly, I and many geeks I know but/bought several when a new model first comes out because they are (were) cheap but electronics wholesalers don’t offer free shipping so it makes sense to buy as many as you think you’ll use (until the next one comes out) in one go to minimize the amortized cost.
I miscalculated and ended up using all of mine, but I have friends that still have some of their stock.
I also installed fiber a while ago after being told "this isn't hard" by Michael Stapelberg's post [1]. Luckily I managed to get by with patch cables; a 30m cables really goes a long way! I even managed to get a duplex LC through a 16mm conduit with a pretty tight 90° bend, but just with lube, quite some force and luck.
I can just recommend going for it if you're curious. Cables and tranceivers really aren't expensive from fs.com (though their sales people may start bothering you).
Edit: And old, used network cards from ebay, of course.
Kudos for pulling those through these small conduits :)
For anyone trying to emulate this: I would very much NOT recommend using 16mm conduits for LC duplex patch cables. It's doable for some short/straight runs, but the connectors like to get stuck in bends. For me it was the final r=3cm bend in a ~20m conduit. I ended up pulling Cat7 into that conduit and routed the fiber through a tree-like network of wider backup conduits.
Since the backup conduit ends up in the wrong corner of the room, I might one day 1. cut the 30m fiber patch cable, 2. pull it into the original 16mm conduit (w/o connectors), replacing the Cat7 and 3. splice the fiber again. Using a "mechanical splice" that should be doable quite cheaply, but I didn't yet get around to learning/practicing that.
> My ISP plans on proving 100gbits in about 2 years so I don’t need pull new fiber if I decide to upgrade.
Jesus. Meanwhile, here in Germany, right in the center of Munich the best I can get is 100/40, because while there is FTTB buildout, the last 20m in the building go through the telephone wiring from the late 80s.
If you really want it, and are willing to pay, you can often negotiate with the building owner to get the last 20m run (up to and including conduit hidden on the outside of the building, etc).
If enough people in the building care, it might even be relatively cheap.
The latter is the problem. Half the people in the building are old, the other half of them are happy with the way it is - I'm the only IT guy. Might be possible to pull fiber through the conduits, but not sure if the Mnet FTTB box can actually give me fiber - I think it's a miniature DSLAM only.
Mnet should be able to CWDM (of they don't just have a fiber to spare for it) you a link on the DSLAM's fiber uplink.
Provided the building owner allows the fiber from the DSLAM to you.
They could also put an Ethernet switch needy to the DSLAM so more people could share one uplink connection.
3M sells fusion-less APC and UPC connectors that you can insert into any (well cut) fiber. I used them to fiber my home too without using a fusion machine.
I was really dubious at first but they work really well
Although this stuff is all mundane to me (network engineer), it was a long, fun read about a layman dipping his toe^W entire leg up to the armpits in networking.
I did this as a small-scale experiment/proof-of-concept earlier this year in my 2 br apartment, preparing for wiring for 10G (and possibly beyond as the years pass) when I move into my house later this.
I went with singlemode fiber. Don't really see the point of multimode. At all. After doing some research I am not sure who it is made for.
Sure, the BiDi SFP+ modules are a bit more expensive but not overly so.
The fiber is cheaper.
It's easier to splice/use field assembly connectors if needed.
Much easier to hide the cables. If you go with 0.9 mm you can even hide it in nooks and crannies under door trim etc.
It is future proof. SMF will always be supported.
Now the only problem is that the Mikrotik CRS routers I want to get are either out of stock or have insanely inflated prices. Seems to still be affected by component shortages. I already have a couple of CRS305 and some CCS610, but I'm going to need more of those now plus a CRS309 to scale it all up.
I think back 15 years or so ago, multimode fibre and optics were quite substantially cheaper. Given that you could still get several hundred metres to over a kilometer with the slower speeds available at the time, it probably was worthwhile. But by the time I got into fibre stuff SMF was not much more expensive, so I've always just used it by default.
I recently had my home wired with cat8. In the back of my head I had wondered if anyone was doing fiber instead. Nice to see this write up of your journey.
Is fiber overall less power hungry than twisted pair or copper? Asking for solar rural aspirations deep into my mind. I like the idea of it not being conductive thus not vulnerable to electrical discharges.
Contrary to what the author and others have written, I would discourage wiring with single mode (SM). For ethernet the cost of modules and patches might not seem so significant, but structured fiber can be used for things other than ethernet, for which SM is vastly more expensive than multimode (MM). Think 4k display port for cinema/follow-me TV, although that is being replaced with SDVoE for 4k@60. However, you do want to use higher quality OM3 or OM4 MM cable (aqua or violet sheath). If required you can carry SM origin signals over OM3/4 by using a mode conditioning patch cable.
In terms of pulling, a lubed cable sock can get through lots of places, and save a ton on termination. Cablers often use a strip of yellow tongue (the nylon tongue on chipboard flooring panels) to poke through a run, and then run a conduit or pull cord. Pre terminated MPO/MPT (which is 12 core) and MPT breakout fans are very cheap in comparison to splicing. It's generally much cheaper to run more fiber than it is to run conduit so you can run more fiber.
> In order to use the SFP+ modules you need the correct PCIe cards on your PC
Are there good SFP+ USB 4 adapters that work on Linux?
One of the bad trends in recent motherboards is minimal number of PCIe slots, which are very clearly focused on GPUs only. So basically there is no room for any extra devices. Their idea seems to be pushing everyone to USB 4 for use cases like that (which supposedly should be able to route even PCIe?).
I'm in the middle of running fiber in my house as well but much less elaborate than OP. I bought a few 100m MPO terminated cables off eBay, a few MPO to LC splitters from FS.com and a few 100m capable optics from eBay and fs.com. All my switches have sfp ports so it was pretty straight forward
I have no experience with terminating connectors directly. There appear to be all kinds available, even ones with gel inside them but they suffer a lot more loss than if I splice a fiber.
Using the splicer gave me the flexibility to splice pig-tail ends or just two fibers together with a small loss.
I have no experience either. But terminating seems easy, for example this guy is terminating singlemode fiber in just a few minutes: https://www.youtube.com/watch?v=uKDYw0jPskc&t=2m35s He uses DINTEK Singlemode ezi-FIBER Connectors which have a low loss of 0.1-0.2 dB: https://www.dintek.com.tw/index.php/singlemode-ezi-fiber-con... Such a low loss seems to be a non-issue for short fiber runs in a single residence, and cheaper than using a fusion splicer.
Your 10G SFP modules have a power budget of 6.2 dB https://www.fs.com/products/11603.html Assuming zero losses at the fusion-spliced pigtail connectors this would be enough for a ~12.4 km fiber link (~0.5 dB attenuation per km). With the connectors I linked, 0.2 dB loss at each end, the power budget is reduced to 5.8 dB which would be enough for a ~11.6 km fiber link.
As you see this makes no difference for a residence where all your links are much shorter than even 0.1 km ! All I'm saying is you had cheaper options than buying a fusion splicer. But I understand you for buying it — it's a cool tool to have fun with, and to get some experience. I would have probably done the same since fiber is also a hobby of mine (though I reiterate I have never built a cable myself) :)
From what I read (never tried and not my area of expertise) for our short runs at home loss isn't that big of an issue, especially for 10km transceivers.
10 Gbps is really a game changer if you have large quantity of data to transfer (multiple terabytes), but 10G cards and equipment are power hungry so I only put out a dangling fiber one in a while to make a backup but the speed is appreciable.
Also I'm still looking for a solution setup a ~200m link, Ethernet and WLAN are not an option, so I'm only left with fiber, 1Gbps would be sufficient, but I won't be able to do it myself because of the price of a splicer. I thought about asking an installer to do the splicing for me, it shouldn't be that expensive, maybe on day.
There are toolless mechanical splices for about a buck each.
Pre-terminated cable also exists, and a single fiber with an LC plug for bidi is small enough to fit through typical conduit.
I have to pass the fiber cable through conduits that are not very large and they are not straight. So I will need to use a pull line. Using pre-terminated fiber will either bock or destroy the connector.
Duplex LC connectors can be split apart and later reassembled. With a bit of fiddling and taping things you can get a cable through a hole that's too small for an assembled duplex.
I'm going to give a try to "fast connectors" that don't need splicing, I ordered a cleaver and a stripper, that's much more affordable than a splicer. Those connector don't seem to be very popular, there is probably a reason. I already have a 1Gbps fiber media converter. I also opted for 4 fibers cable, I will try a temporary 10Gbps P2P link along the 1Gbps. Fast connectors may not be ideal with a multi-fiber cable, but I will try to make it work. The total cost seem fair.
This is likely not useful in most west european homes (even gigabit connections are unheard of in the uk, 10 megs being advertised as “superfast”, let alone tens of gigs). But still a fun project!
The uk is behind most of europe. Even east eu countries have fiber to a significant chunk of the population now, and west is doing great. In france i get 5Gb/1.2Gb for 49 euros. Switzerland is even better. When i sold routers to businesses they would always be gigabit fiber available, so much that i had stopped checking.
> Even east eu countries have fiber to a significant chunk of the population now
Yes, “even” the barbarians of east europe have it. In romania you get around 10 gbps per second for roughly 10 euros a month from what i hear, and pretty much everyone has fiber. I believe Lithuania, Estonia and Latvia are doing great as well.
Peering outside of Romania might be questionable. I visited a few months ago and got 500Mbps within the country and just 50Mbps to my server in The Netherlands.
I tried running speed tests to other datacenters in NL and had similar issues across the board. It was probably just an isolated incident that weekend, but that was my experience regardless. When I did speed tests domestically, I had 500Mbps both up and down.
They'll eventually be dragged into the future by requirement, 1GbE is already standard on most equipment, with 2.5GbE and 1+gbit wifi becoming more common on consumer equipment. The increasing reliance of on-demand delivery of various forms of content, from multimedia to gaming, is already pushing 100Mbps connections to where 10Mbps connections were less than 5 years ago.
I think its worth choosing fiber in a new deployment, at least at the distribution level (between floors, or zones of the home). Most new computers come with ultra fast solid state storage, so even between systems on a local network, 10GbE speeds could be useful.
With the way the price of copper is going, and how much less efficient rj45/copper chipsets are compared to optical, I wouldn't be surprised if we see a shift to optical as a replacement to rj45 in the next decade.
At a time when more internet developed countries are moving towards 10gb connections, the uk is moving towards 1gb. Instead of playing catch it should lead. Now if only there were competent people in power.
Fast local network is useful even without fast uplink, e.g. for accessing data on local server. 1 Gbps is still several times slower than disk speeds, so going to 10 Gbps seems like reasonable future-proof config.
This is not an accurate statement. I'm paying GBP25/mo for 1Gbit symmetrical in London suburbs. Same provider offers up to 3Gbit symmetrical for non-business. And I'm not in some fancy new build- the fibre is run from the existing street telegraph pole to the front of the house.
Yes it's not necessarily nationwide but "unheard of" is way off, before you even consider Virgin cover a sizeable chunk of the country and their DOCSIS upgrade means they can push gigabit now.
About 70% of UK homes have access to a gigabit connection through either fibre or the Virgin Media HFC network. FTTP coverage is about 40% nationwide. It’s not that rare.
Respectfully, I find the 70% figure for access to a gigabit line extremely hard to believe.
When I lived in central London (tube zone 2), all that was available at my address was a wet-noodle quality 20 Mbps DSL connection. No BT FTTP, no Virgin, no Hyperoptic, no G.Network. This was in early 2022.
I now live in Switzerland and have a 25 Gbps FTTP connection, paying about 30% more than I paid in London for the DSL line.
It's a completely different world connectivity-wise.
FWIW last time I checked, central London did indeed have slow connectivity, but it didn't extrapolate to the rest of the UK.
No idea why the center of the capital was equivalent to a remote small town in this regard.
I agree, that's very counterintuitive. You'd think the economics would generally be in London's favour, given the high population density. (Though digging up the streets might be harder and therefore more expensive?)
Well i clearly couldnt find such homes and believe me i looked. I was happy to have found homes with proper structural integrity let alone high speed internet.
That's because the UK is way behind mainland Europe in this regard. My previous house in a very busy area of London had 80mbps max...
However in a small town in Switzerland we have 25Gbps available. The UK could be in this position, but Thatcher happened unfortunately, and the project was canned.
You'd think simplex short boot LC would be able to get through small conduit. Although you can always stick the fusion splicer on eBay, it's not like it's worthless now.
Wiring for me has been a waste of time. I spent a good amount of time wiring the house with cat 6 cabling. I had our TV on a hard wired network which felt great, no drop outs, no wifi noise. I then bought a roku box and left on wifi for a while. Its still on wifi, works great and kept it there. Family mostly uses wifi now, no PCs using the hard-won ethernet jacks.
I ran fiber between the network core and my bedroom (which housed my home lab) at my parents old house circa 2004, because I got access to cheap fiber and a fusion splicer through school at the time. I’ve never run fiber since because the usefulness of POE outweighs any benefits fiber might bring and STP is good enough in home environments.
25Gb Internet? Bandwidth in the US sucks! But I have to admit I have trouble saturating 35Mb, and I'd happily pay half as much for 10Mb or 5Mb if it were available. Also, tablets would be pretty annoying without wifi.
Thanks- I rarely consider upgrading my home systems until I think I have a good understanding of all the details and I see that fiber is still a mess (have to choose single mode versus multi mode and a number of other details).
Thank you and the others who also submitted it as originally this post was flagged. Probably because I changed the title after submitting it to "Show HN" but I am unsure.
I am glad now that I spent the time to put it on github.io and not on some old server of mine that would have been hugged to death.
You pay the SFP module tax on every port for that model, and also have a loud fan. You also probably can't put a copper SFP module in every cage without overheating.
That will let you receive 10Gbps fiber, and maybe route 3Gbps or so. No fans. No SFP module tax. And PoE on every port.
Mikrotik is definitely missing a product that has maybe 1 SFP28 port, 1 SFP+ port, 4 10GBps ports and 5 2.5Gbps ports. That would let you take in 25Gbps fiber, have a 10Gbps fiber run to a switch someplace, and run 10Gbps ethernet as needed.
Mikrotik has plenty of switches with those specs, but they won't be able to route at line speed. If you can arrange your network to be primarily switched rather than routed (i.e. all the high speed hosts have their own static IP addresses, and the rest are behind a slower router), then you can do that.
Otherwise, you can get a powerful router with 2 SFP+ ports and link an SFP+ switch for the rest.
Mikrotik has some very affordable & capable routers under $1000 with SPF or SPF+.
Depends on how many ports, but they'd start with RB5009UG+S+IN which has a suggested retail price of $219.
I wonder what is use case for this network. At my home I only use optics for long run, more for galvanic isolation, as electric charge could destroy electronics. All servers are concentrated in single room. Clients use miniPCs. Rest of the house is wired with 2.5Gb ethernet and wifi6, anything faster and expenses grow exponentially.
25Gbps would be hard to saturate with an NVME SSD. Even if you like to mirror hard disks to remote locations all the time, you can just RSYNC, and then your needs are limited to the rate of change.
I'm getting a 10Gbps from the same provider, and even that is not likely to saturate, and I'll only have 1Gbps CAT6 links. The main benefit is that the clients are absolutely independent, i.e. downloading a Linux image can't affect someone's video conference, since each client can only pull 1Gbps of 10Gbps total.
Well that multiple client fairness gets solved by using something like CAKE. There are youtube demos doing exactly this even with links limited to like a few meg.
The author's ISP delivers up to 25/25Gbps connections (and 10gbps connections for the same price as 1gbps connections, about $65 dollars per month converted) so if he is planning on using such a high bandwidth uplink, I don't think ethernet makes a lot of sense.
If I had a 10gbps uplink for that price, I'd certainly look into getting more out of my network than just standard ethernet. 10gbit ethernet at least, though fiber may be easier to install depending on the size and layout of the house.
I have 1gbps fiber and it's extremely rare for me to come across a server that supports anything close to that bandwidth. At 10gbps you're able to upload 500GB/hour. For home use this seems extremely unlikely, even if you're using this as part of your job. Even my 4K surveillence cameras only require 20mbps each, that's 50 4k cameras to saturate a 1gbps line.
I think of >1 GBps speeds as something that serves burst rather than streaming needs.
The use case isn't "I need it to watch youtube", but "I want to be able to restore from backup in hours rather than days", or "I want to play the latest Doom today and not tomorrow".
Eg, say you're backing up your data to a remote site. Great idea, but what if you need a restore, how long will that take? Downloading say, 100 TB on a 1 Gbps connection will take you more than a week.
That would only works if the other end can push 25gbps as well. I wonder what's the maximum throughput of various cloud storage services commonly used for off site backups (S3, B2, etc). Would they artificially limit the max bandwidth or allow you to go as fast as possible?
I've had trouble saturating my 1 Gbps connection in Sweden. I did tests with B2 and Wasabi roughly every quarter for a couple of years trying to see if it was feasible to move some data hoarding activities there, and never got more than ≈100 Mbps when downloading from them.
Don't know if it's still the case, or if my ISP was to blame (or just being in EU/Sweden).
On the other hand, I don't have a problem maxing out 1 Gbps when downloading both metaphorical and actual linux iso's. A lot of the microsoft stuff is really fast as well, wouldn't be surprised if they could saturate ≈ 10 Gbps.
Have to think about the future though when you’re dealing with running stuff in walls. Think about what would’ve been acceptable throughput 10 years ago. If you built your home network around that, you’d probably be kicking yourself today.
Even CAT5e supports 10gbps in your typical house (you'll want CAT6 for longer distances though), and 10gbps is likely future proof for the next 25 years (which is crazy because CAT5e came out in 1999).
I think one major use case in this particular scenario is that the author has one 25 Gbit/s and one 10 Gbit/s internet service. If you want to make the most use of this fibre is the only way to go.
I do data processing and can saturate similar bandwidth. But 10Gbps goes to my home server room.
There is no need to have 10Gb in bedroom, such tech produces a lot of noise and heat. Maybe Stadia or similar video streaming could use such bandwidth. But this looks more like wiring for residential building with multiple flats, or office building. Or like some sort of tech flex.
…until later in the home’s life when that bedroom is no longer a bedroom, for whatever reason. If you’re gonna do a project like this you might as well take it to its logical conclusion.
It’s the same kind of advice I’d give to anyone who buys e.g. an iPad Pro - you might not think you need a cellular model right now, but that one time you do need it two years from now you’re gonna be very glad that you paid the extra ~$100.
That’s some very hypothetical talk. The master bedroom in a typical home is likely and forever going to remain a bedroom, at least if there is sufficient other space available in the home. If you buy a couple of iPads (you and your spouse or kids or whatever) and upgrade them every x years, springing for an extra $100 on each one every time adds up. You never need the cellular module, though it might be nice to have since there are always other, less convenient options (hotspot on phone, standalone hotspot, public Wi-Fi, etc).
You cannot take your mantra to its logical conclusion and apply it to everything, everywhere, at least not unless cost is absolutely a non-issue for you.
> there are always other, less convenient options (hotspot on phone, standalone hotspot, public Wi-Fi…
Until you’re driving, hypothetically, through rural Arkansas, where there's no Starbucks and your iPhone has juuuuuust enough signal strength to receive the WhatsApp telling you that you need to turn around a couple of slides but not enough bandwidth to download the deck.
Your iPad, though, for reasons known only to the black magic gods of RF design does have a stable-enough LTE connection and it saves your bacon with the large corporate client that dragged you to razorback country in the first place.
Hypothetically.
> You cannot take your mantra to its logical conclusion and apply it to everything, everywhere
There’s a reason I wrote iPad Pro. It’s not my mantra - it’s advice that applies to specific kinds of people in specific kinds of situations. The kind of people who have the means and motivation to run fiber throughout their house, for example.
Hindsight is 20/20. There’s a reason why insurance policies exist. You can (and should) save money skipping comprehensive insurance on a beater. Unless you are going to get into a car accident found to be your fault in the first year of the policy, in which case it would have made more financial sense to get that comprehensive coverage. But that logic doesn’t hold.
10Gb normal CAT whatever Ethernet works but the SPF adapters get quite hot.
I have a 10Gb backbone for my “servers” and then a single 10Gb Ethernet cable to my main Mac- works well enough for now and probably not worth upgrading until I have internet beyond 10Gb.
That depends on the food. My parents mostly grew fruit, and berries especially are very expensive at the supermarket.
The other things they grew were also equal or better to the most expensive produce at the most luxury supermarket, like tomatoes that tasted of tomato rather than the watery, red golf balls they sell in Asda.
There are other benefits of fiber, such as transmitting non-Ethernet signals such as USB, DisplayPort, HDMI, etc. You can run these things over Ethernet, but they are really shoddy and unreliable.
It allows you to centralize more things. For example, you could just have a monitor, keyboard, mouse, and a hub in a room for a computer, where the computer is actually somewhere else (such as a server room).
While I like the idea I do not like much fiber or to be more precise their live stuff, since they degrade far faster than copper. A copper switch might work well for 10+ years, and since at home we normally not need such extreme performances and these days 2.5GBE on cat 8.1 copper are relatively cheap and common...
1. Just use UPC connectors everywhere unless you have really good reasons not to.
2. Use single mode fiber (SMF) everywhere, including internal wiring. SMF is infinitely scalable for future needs, multimode (MMF) is not. These days the cost differences for cabling and optics is negligible. In fact, SMF cabling is often cheaper than MMF. And no, you generally won’t have to worry about attenuating signal.
Also, in my extensive experience, FS.com optics modules are typically better manufactured and more reliable than OEM modules.
I resisted the "cheap chinese" modules for a long, long time, but then discovered that most of my uplink carriers used them.
It's been years now. I've had many Cisco OEM, Dell OEM, and a couple from another 3rd party manufacturer go bad, but I have yet for one FS.com SFP to go bad. I've got several hundred deployed in a wide variety of environments, from great (cool, clean areas) to bad (hot, dirty areas).