Posted by alienchow 1/11/2026
One random guess other than gases of various sorts: If you have the lights on a lot, it's possible you have lighting that is not properly uv filtered somehow.
(It theoretically could be lots of heat/humidity vs certain types of cable jackets, but that would be obviously noticeable most of the time)
The fiber and even the armor itself looks fine in your pictures and I bet the error rate is zero. Outside of bending, i can't see how you could damage the fiber when the armor is in perfect shape. That armor will unravel if you really damaged it. The speed thing is not how fiber works. It's hard to get marginal link enough to generate retries that degrade your speed by 10%. Most of the time either you have full speed link, no link, or so many errors speed is zero. The optics almost always have rx/TX signal strength DDM you can look at
Also did you say you direct buried it in cement? If so it's not rated for that. Direct burial and concrete tight/safe are not the same thing at all, not the least of reasons being concrete is highly alkaline (ph12-13) when poured
The second possibility is that I keep the leftover wall paints (Nippon Paint Vinilex 5000) in the same room and have noticed that much of the solvents have evaporated. It is possible that the solvents in the air might have caused the cable to fail in 3 years. It would explain why the other ends that aren't exposed to the air inside the bomb shelter aren't falling apart.
Some other learnings from this. Buried cabling should always be permanently fixed and attached to a patch panel instead of dangling in the open. That was the original plan but I figured it wouldn’t be an issue. I was wrong. Always measure exact length of buried fibre cabling as they aren’t meant to be stored in loops.
This is awful advice I would discard immediately. It's poor practice and against code.
When pulling cable, especially fiber, the ends of the cable should be able to reach the fathest corner of the room. Excess cable should be in a service loop, properly secured to a wall, and terminated on a patch panel. Both ends of he cable should follow this rule. That means you're typically pulling cable that's 15m longer or more, depending on the room and configuration.
NEVER buy and pull cable that is the exact size. The cable literally comes from the factory looped up, it's designed to be looped (watch bend radius).
if you're using 25gbps/sfp28 transceivers, you probably have FEC. if so, you probably have both correctable and uncorrectable error counters to look at.
That said military spec stuff is actually generally a good sign that something is of higher quality than random off the shelf garbage but only if you know there's a specific spec you want it to work with. And most of the time you aren't even necessarily looking for a MIL-STD (standard) but rather a MIL-PRF (performance rating/spec).
So like if something is "MIL-SPEC" run. But if you see say a spool of fiber that is "MIL-STD-1678 compliant" and more importantly "MIL-PRF-49291 compliant" and "MIL-PRF-85054 compliant", that's probably a really good sign that it'll do its job. The former PRF documenting perf requirements for the fiber itself and the latter PRF the cabling/sheath's corrosion and deterioration resistance.
It's the military so odds are it'll probably cost extra for that and it'll still kinda suck but it'll suck in exactly the way they promised.
* https://www.navsea.navy.mil/Home/Warfare-Centers/NSWC-Dahlgr...
* https://landandmaritimeapps.dla.mil/programs/MilSpec/ListDoc...
Nothing sucks more than having the engine crap out, 150Km offshore, because your fuel injection system got corroded.
Claiming to conform to a more specific product or process standard would be more specific fraud.
But in general though "military grade" is a red flag for shitty marketing.
Example: pop tarts are military grade! [1]
Though their commercial packaging is likely not.
https://www.dla.mil/Portals/104/Documents/TroopSupport/Subsi...
https://en.wikipedia.org/wiki/United_States_Military_Standar...
It's just a weak pander to people's weak egos. Freedumb, if you will.
On the other hand, when a product is designed and manufactured to sell to a military, it's going to be expensive, and that extra cost isn't going to quality or capability, it's going to compliance. You're more than likely paying extra to get something using some old and outdated technology, that includes paperwork to prove that it's only built using the approved old and outdated technology.
Common mass produced products manufacturers have incentives to not mess-up too badly: recalls or warranties on such scales are a nightmare.
With military contracts, its a paid maintenance opportunity.
This
In the military, military grade is a synonym for crap
Is one supposed to send it back for a refund and order the much thinner, less-durable cable? Or is perhaps the landscape not as black-and-white as your "this is automatically snake-oil"?
Anyone who thinks the triggers listed as MIL-SPEC from, say, Geissele here:
https://geissele.com/triggers.html
aren't totally fine is out of his mind. They're amazing triggers, widely used and loved.
And they don't say which specs its passing (at least not on the main page): it's just MIL-SPEC.
As a sidenote my very best laptop passes https://en.wikipedia.org/wiki/MIL-STD-810 but most people will just say it's "military grade" or "MIL-SPEC".
Guess what? Its screen never broke overnight like the one of my MacBook M1 Air did (the infamous "bendgate").
I can bend my LG Gram's screen and it's keeps working fine. I can let it drop. Friend who sold it to me stepped on it when he woke up once.
There's a very big difference between saying: "There are shady vendors" and saying "Military specs do not exists and it's impossible for consumers to buy items passing military specifications".
Yes, there are dishonest vendors.
Yes, military specs do exist.
And, yes, it's possible for consumers to buy products passing (and even surpassing) actual military specs.
"Military grade" is generally shit. It's built down to a price, manufactured the cheapest possible way, so they can get the lowest possible tender submitted. Bonus prize if the manufacturer is owned by either someone already in government, or with close ties to someone in government.
The only "military grade" devices I own are some woefully unsuccessful radios, which failed in the market because they were actually good - easy to use, reliable, and easy to repair - which made them about 5% more expensive than the cheapest option which was made by a company part-owned by the government and part-owned by someone who donates heavily to the Conservatives.
Military grade snake oil?
When I see "military grade" I assume overpriced $30,000 hammer
I have a similar problem on my car where the 12v wiring is disintegrating like this because the manufacturer tried to switch to a more environmentally friendly wiring. Now the wire jackets turn to dust at the slightest touch or if they vibrate too much. I'm forever tracking down intermittent shorts in the wiring harness.
Honestly, this writeup is… weird? Dude doesn’t know you can terminate fibre at home with like $50 of gear?
I had the fucking fox attack a freshly laid 500 meter line, literally the day before I was going to stuff it in conduit and bury it. Didn’t just break the fibre, she (I know this fox, well) chomped it into pieces, hauled on the exposed Kevlar, generally had a party.
Did I despair? Did I launch a baby complete with bathwater into the sun?
No. I bought a cleaver, some alcohol wipes, some stripping pliers and a whole bunch of mechanical terminators.
Needn’t have worried. Repaired it, outdoors, first attempt, in the rain, and have since buried it - no problems five months on.
Unfortunately I can't easily dig the cable out and bury it again in this case. I'll have to figure out how to pull a new cable using the existing cable through the PVC conduits as the cable shares a larger conduit with multiple other fibre and Ethernet cables. The whole project was orchestrated remotely in a different timezone with me giving the electricians instructions over WhatsApp photos and audio recordings, so that limited what I could realistically control onsite back then. Often the contractors would proceed with a do first ask questions later approach while I was still asleep. The networking project was holding up the entire home renovation so everything was learnt and planned in a short amount of time.
AFAIK fibre splicing and terminating tools are very expensive. Do point me in the right direction for the $50 tools and I could go get some and DIY.
They're more like $600 expensive than $6000 expensive these days. For very low budget, you could go with a mechanical (aerobic) splice; it's more loss, less robust and takes up more space, but doesn't require a fusion splicer.
- secure a string to the old cable
- pull the cable out the other end, pulling the string through
- secure the string to the new cable
- pull the string out the other end, pulling the new cable into position
At this point, multiple. Just in case.
This gives you a piece of string in the conduit run to be able to pull through the next thing days/weeks/years later.
Does someone have a recommendation for a specific material the string should be made of?
You basically want something that is slippery and will tend to not get stuck. I have used Dacron fishing line, but that is mostly because I had a bunch of it laying around.
I wanted to reach you regarding your comment here: https://news.ycombinator.com/item?id=44547866 Would you be willing to maybe elaborate on the problems caused - I've planned to adopt Miro Sameks for an application? DM me via my about me, if interested. Would be very thankful.
I have issues with the PVC-jacketed cables under the bonnet of my nearly-30-year-old Landrover, where the plasticiser has been baked out of the insulation and they've gone brittle. Worst affected are the wires to the fuel injectors and the lambda sensors, presumably because the former are at the top of the engine and get reflected heat off the bonnet, and the latter because they're near the literally red-hot exhaust downpipes.
That's okay for an old vehicle that you'd expect to repair, though.
I've seen the same problem in three-year-old Toyotas, and that is Just Not On.
The bio-oil plasticizers also migrate out more quickly in thermal cycling than the old dead dinosaurs approach. Hilariously, when I asked my mechanic about getting an M5, he laughed and explained that the radiator components are known to turn brittle and crack after 5-6 years because of this.
(I don't envy automotive folks. The stuff they have to deal with is next level.)
Ugly AF but artisanal and robust.
( Just avoid the original third gen versions from late 1980s (?) that offshored the assembly )
2012 Bus configuration: https://www.youtube.com/watch?v=fchq6F-gKkQ
Another thing that should not have happened is installing the cable in loops in this way: any 'building' or 'underground' type cable needs to be of the exact length required at the demarcation point, fastened properly to prevent movement and terminated on a proper patch panel (can be a one-port box-type thingy for small setups), from where you use regular patch cords to connect your equipment.
(Loops are definitely allowed though, but that use case is mostly for aerial fiber to enable repair splices, and there are some very specific bend-radius and strain relief requirements, which, again should be spelled out in the cable data sheet)
How exact is exact? :-) I once had to reterminate some fiber that was cut and terminated to exact length, which means there was literally two centimeters from the wall to the connector. I literally had to squeeze the fiber splicer up against the wall to have a chance at splicing on new pigtails, but I had two mis-cuts and I was hosed. :-)
This hasn't been my experience with fiber entrance cables terminated by ILECs, Spectrum, and Lumen. They typically leave a significant service loop bound to the cable ladder or backer board-- usually 15-20 feet.
Definitely learnt it the hard way this time. You're right that buried cables should be exact in length and fastened to a patch panel. I'll probably look into better conduit design as well for the next time (in 15 years?). Having shared conduits means I would risk damaging other cables if I tried to pull a new cable through.
Sharing/in-place-repurposing conduit is not something I'd recommend, but if you must, leave a few dummy cables (a.k.a. 'pieces of string') on the initial install...
But yes, agreed, a lot of "Er... why would you do it like that?" bits.
Cables for direct burial only like to be bent once or twice, and then only gently. Anything else may very well break the armor (whether plastic or metal), after which all bets are off.
Still, for the outer jacket to become brittle to the extent described, something else is required, which may very well turn out to be "shoddy manufacturing"...
This is awful advice I would discard immediately. It's poor practice and against code.
When pulling cable, especially fiber, the ends of the cable should be able to reach the fathest corner of the room. Excess cable should be in a service loop, properly secured to a wall, and terminated on a patch panel. Both ends of the cable should follow this rule. That means you're typically pulling cable that's 15m longer or more, depending on the room and configuration.
NEVER buy and pull cable that is the exact size. The cable literally comes from the factory looped up, it's designed to be looped (watch bend radius).
>Loops are definitely allowed though, but that use case is mostly for aerial fiber to enable repair splices
Again, awful advice that's against code. Underground fiber must have service loops at both ends, and must be terminated to a patch panel.
While some show brittleless, more plstics goes gooey and tarry - especially some ABS coating that makes the material more grippable,like computer mice or binoculars.
I bought a big spool of 6 strand Corning stuff a long time ago for various projects, the cost and diameter don't increase much to add some protection lines even if you never imagine using them they can save you a re-pull if you bugger something up in construction.
Honestly, fibre, even unarmoured with just a standard Kevlar & HDPE sleeve is hardy stuff. When I first started mucking around with it a few years ago I was like “don’t breathe on it too hard”, now I’m like “tie the fibre in a knot on the bullbar and pull it with the truck”.
To be fair, it also got a lot better in the last 20–30 years. In particular, we now have bend-insensitive fiber for the last mile (G.657.A1/G.657.A2) and in general, we just figured out how to make it more robust.
Speedtests for 10G are complicated and will show low numbers because of all the different TCP parameters and schedulers. Sometimes because peering links of your ISP or the speedtest providers are saturated.
My internet network would would test at theoretical limits with proper iperf2 settings. I tried using public iperf servers but wasn't successful.