Learning how to fix a router with no internet is mostly a process of elimination done in the right order, because ‘no internet’ has only a handful of causes: your provider is down, the modem has lost sync, the modem and router have fallen out of sync with each other, a cable has failed, or — less often — a router setting or the router itself is at fault. This guide works through them safely and in sequence: read the lights, power-cycle in the correct order, test with a cable to isolate the faulty half, then check settings. Do not factory reset anything until the later steps; resets erase the very settings you may need, and most no-internet faults are cured long before that point.
If the internet works but devices keep dropping off Wi-Fi, your guide is How to Fix a Router that Keeps Disconnecting instead — a different fault with a different fix list.
Disclosure: As an Amazon Associate, ilcastellomenu.com earns from qualifying purchases. Links to Amazon on this page are affiliate links; they cost you nothing extra. Listing details such as price and availability change often, so check the current Amazon listing before you buy.
How to fix a router with no internet: read the lights first
The modem and router are telling you where the fault lives; look before you reset the evidence away. On the modem (or fibre ONT): a solid online/sync light means it is connected to your provider; a blinking or red one means it is not. On the router: its internet/WAN light shows whether it can see the modem’s connection — green or white is good, red or amber is not. Three combinations cover nearly everything:
| Modem light | Router internet light | Likely fault |
|---|---|---|
| Off / red / blinking sync | Anything | Line, provider outage, or modem — start at Step 1, then Step 4 |
| Solid online | Red / amber | Between modem and router, or router settings — Steps 1–3 and 5 |
| Solid online | Solid / green | Internet is reaching the router; the problem is Wi-Fi or one device — Step 3 |
In this guide
Step 1: Power-cycle in the correct order
Unplug both the modem and the router. Wait at least thirty seconds — longer if the modem has a battery backup, which must also be removed or the modem never truly turns off. Plug the modem in first and wait until its online light is completely solid; this can take up to five minutes and cannot be rushed. Then plug in the router and give it two to three minutes. This single procedure — the same one detailed in How to Restart a Router — cures the majority of no-internet events, because the modem and router renegotiate a fresh connection. Test before moving on.
Step 2: Check the cables
Confirm the Ethernet cable runs from the modem to the router’s WAN/Internet port — not to one of the numbered LAN ports, a mix-up that happens after any furniture move or cleaning. Reseat both ends until they click. If the fault persists, swap in a different Ethernet cable; cables fail quietly, especially ones that have been kinked behind furniture or chewed by pets. Also confirm the incoming line (coax, DSL or fibre) is firmly seated at the modem and the wall, finger-tight for coax. Never open the modem, ONT or router case — there is nothing user-serviceable inside, and these are mains-powered devices.
Step 3: Isolate with a wired test
Plug a laptop directly into the router’s LAN port with a cable and test. If wired internet works, your connection and router core are fine and the fault is Wi-Fi only: try forgetting and rejoining the network on an affected device, and check whether one band has vanished (a sign of a band-specific fault or a bad update — see Step 5). If wired through the router fails too, bypass the router: connect the laptop directly to the modem, power-cycle the modem again with only the laptop attached, and test. Working direct-to-modem but not through the router convicts the router or its internet settings (Step 5). Working nowhere convicts the line or modem (Step 4). This isolation logic is the same modem-versus-router reasoning explained in Modem vs Router.
Step 4: Rule out the provider
Check for an outage before assuming your equipment is guilty: your provider’s app or status page from your phone (on mobile data), their recorded line-test if offered, or neighbours on the same provider. If the modem never achieves a solid online light after a correct power cycle and cable check, and there is no declared outage, the fault is the line or modem — report it, describing the light behaviour and what you have already tried; that detail gets you past the scripted first line of support. If you own the modem, confirm it is still on the provider’s approved list and matched to your plan, per the provider-specific guidance in Best Routers for Xfinity or your own provider’s equivalent.
Step 5: Check the router’s internet settings
If the modem is healthy and the wired bypass test worked, sign in to the router’s admin page and look at its internet/WAN status. It should show an obtained IP address; a row of zeros means the router is not being given a connection. Check that the connection type matches your service — most cable and fibre connections use automatic/DHCP; DSL and some fibre plans use PPPoE with the username and password from your provider’s paperwork, and a typo there produces exactly this fault. If you recently replaced the router, some providers bind the connection to the old device’s MAC address: either wait (it often clears), use the router’s MAC-clone option with the old address, or ask the provider to re-register. While in the settings, confirm the firmware is current via How to Update Firmware on a Router once the connection returns.
Step 6: The factory reset — last resort, done properly
Only if everything above fails and the router itself is convicted: back up or photograph your settings if the interface allows, then factory reset using the recessed button (hold for the duration the manual specifies, usually around ten seconds, with power on). The router returns to its sticker defaults, and you set it up again following How to Set Up a Router and re-harden it with the security checklist in the companion guide. If a freshly reset, correctly configured router still cannot hold a connection that works direct-to-modem, the router hardware has failed; replacement shortlists are in our home-router shortlists.
Prevention
- Keep the provider’s app installed — outage checks take seconds when the internet is down and tempers are short.
- Label the cables and photograph the working setup; future faults get diagnosed against a known-good picture.
- Automatic firmware updates on, and a spare Ethernet cable in the drawer.
- If outages are frequent and you work from home, a router with automatic failover to a mobile connection is the structural fix — see the work-from-home router shortlist.
Variations by connection type
Fibre (ONT on the wall)
Treat the ONT as the modem in every step above. Its key light is the optical one: red or off means the fibre path itself is broken or the ONT has failed — a provider fault you cannot influence, so report it early rather than after an hour of router experiments. A healthy ONT with a router showing no connection points at the router’s WAN settings: some fibre services need PPPoE credentials or a VLAN ID, and those live in the provider’s welcome pack.
DSL
DSL sync is the modem light that matters, and DSL has two household-specific gremlins: the line must run through its filter/splitter correctly, and adding or moving a phone, alarm or extension without a filter can kill sync outright. If the outage began when something was plugged in anywhere in the house, unplug it and retest. Long internal phone wiring degrades DSL; the modem belongs at the master socket for diagnosis even if it normally lives elsewhere.
Cable
Cable modems are sensitive to signal levels that shift when splitters are added or coax is damaged. If the outage followed moving the modem to a different coax point or adding a splitter for a TV, undo that change for the test. A modem that syncs at one socket and not another has told you the house wiring, not the internet, is the story.
After storms and power cuts
A disproportionate share of ‘sudden’ no-internet faults follow weather or a power event, and they have their own pattern: equipment comes back in the wrong order, a modem hangs half-booted, or — the unlucky case — a surge has damaged the modem, the router, or the Ethernet port between them while leaving everything apparently alive. Work it fresh from Step 1 rather than assuming yesterday’s configuration is today’s reality: full ordered power cycle, then the cable swap in Step 2 gains importance (a damaged cable or port is now a prime suspect — try a different LAN port on the router for the wired test as well). If the modem never syncs after a storm and neighbours are also out, it is the provider’s network, and your job reduces to reporting and waiting. Households in outage-prone areas can soften all of this with a surge protector for the networking shelf and, for the work-critical, a small battery backup that rides out the flickers that cause half these faults.
Talking to provider support so it actually helps
Support calls go better with a script of facts. Before dialling, have ready: your account details, the modem’s light states right now, whether a wired connection direct to the modem works, what you have already power-cycled and in what order, and when it last worked. State the isolation result early — ‘no sync light after two power cycles, cables reseated’ moves you past the restart script to line tests. If they send an engineer, ask what was found and note it; recurrent line faults have a history, and your note of the last visit’s finding speeds the next one. And if the fault turns out to be your router after all, accept the verdict gracefully — the isolation test in this guide exists precisely so that you can be sure before you spend money, whichever way the evidence points.
The ‘it worked yesterday’ fallacy
When a connection dies untouched, the mind insists nothing changed — but the network’s environment changes constantly without asking. Providers push gateway firmware overnight and occasionally change authentication details. A neighbour’s new router lands on your channel (a congestion fault, but it can present as near-no-internet at peak). A tree grows into a fixed-wireless path. The cat sleeps on the warm modem and slowly cooks it. Listing this is not whimsy: it licenses you to run the full sequence in this guide even when ‘nobody touched anything’, because something always touched something. The sequence is cheap — fifteen minutes — and it does not care whether the change was yours, the provider’s or the cat’s.
When no-internet becomes a pattern: the escalation ladder
A single outage is an event; three in a month is a case. Climb deliberately. First rung: your own log — dates, times, light states, whether the wired bypass worked. Second: the provider’s line test and a fault reference number; insist on the reference, because it converts your next call from a new story into a continuing one. Third: equipment substitution — a provider-supplied replacement modem, or your router swapped for a known-good spare for a week; substitution is the most decisive diagnostic in all of networking, and providers will often supply the modem half free. Fourth: an engineer visit, armed with the log. Fifth, if a provider cannot or will not fix a demonstrably faulty service: the formal complaints route and, ultimately, a different provider or connection type if your area offers one. Most cases resolve by rung three. The ladder’s purpose is to stop the two failure modes of household troubleshooting — giving up at rung one, and buying new routers at random, which is rung zero with a credit card.
A five-minute monthly habit that prevents the worst version
The worst no-internet is the one during the meeting, the exam, the deadline. Prevention is mostly rehearsal: once, while everything works, confirm you know where the modem and router are, which lights mean healthy, where the spare Ethernet cable lives, and that the provider’s app is installed and signed in on your phone. Photograph the healthy lights. That is the whole habit, plus the quarterly firmware glance this series recommends elsewhere. Households that have done it experience an outage as a procedure; households that have not experience it as an emergency with a router manual. The fifteen minutes in this guide are the same fifteen minutes either way — the only choice is whether they happen calmly, once, in advance.
The ten-minute version for the impatient
Everything in this guide compresses into a drill worth rehearsing once, calmly, before you need it under pressure. Glance at the modem lights: online light solid or not — that single observation routes the whole diagnosis. If not solid: power-cycle the modem properly, check its incoming cable, then call the provider if it will not sync; your work is done and the fault is theirs. If solid: power-cycle the router, confirm its internet light, and test one wired device. If wired works, the fault is Wi-Fi only — reboot, rejoin, update. If wired fails through the router but works direct from the modem, the router’s settings or hardware are guilty — credentials, connection type, or replacement. Memorise the shape rather than the details: lights, order, isolation. The full guide exists for the branches; the drill exists for the moment, and the moment is usually when someone is asking you, from the other room, how long the internet will be down.
Households that cannot afford downtime
For some homes ‘no internet’ is not an inconvenience but a risk — remote medical monitoring, a carer alarm, a business run from the kitchen table. If that is your household, the fix is structural rather than diagnostic, and it is bought before the outage, not during it. A router with automatic failover switches to a mobile-data connection when the main line dies and switches back when it recovers; the household notices a pause, not an outage. A battery backup keeps modem and router alive through the power flickers that cause a large share of domestic dropouts. And a second, independent connection — even a modest one — is the only true cure for a provider-side failure. None of this replaces the diagnostic sequence when things break; it ensures that when they break, the house keeps breathing while you run it.
After it comes back: the five-minute post-mortem
When the connection returns, resist the understandable urge to simply move on. Spend five minutes while the evidence is fresh: note what the lights showed, what you did, and what apparently fixed it, in the household log or a note on the fridge. Check the router’s uptime and connection statistics if it keeps them — a connection that renegotiated forty times last week tells a story the single outage did not. Confirm the firmware is current, because an outage is often the first symptom of the buggy version an update already fixes. And if the fix was ‘the provider did something’, note the fault reference they gave you. The next outage — there is always a next one — starts from this page rather than from zero, and a household with a post-mortem habit finds its faults getting shorter, rarer and considerably less dramatic.
The short version
Read the lights, power-cycle modem-first, check and swap cables, isolate with a wired test, rule out the provider, then — and only then — settings and reset. Followed in order, this sequence finds the guilty half of the system within fifteen minutes in almost every case, without erasing anything you will regret.
Gear mentioned in this guide
You do not need new hardware for most of the steps in this guide, but the products below are the ones referenced above. Each was verified against Amazon’s Creators API product data for this site, so the link matches the exact model named.

ARRIS Surfboard G36 DOCSIS 3.1 Multi-Gig Modem & Wi-Fi 6 Router
The ARRIS Surfboard G36 combines modem and router in one DOCSIS 3.1 unit for cable households — relevant if your diagnosis convicts an ageing rented gateway and your provider approves this model for your plan (verify on their list first).
The impostor outage: when the internet is fine and only names are broken
There is a failure that wears this article’s title as a disguise, and catching it early saves the entire reset routine. Your devices report “connected, no internet”, every website fails — yet the line is alive, the modem is synced, and the router holds its connection. The broken piece is name resolution: the system that translates website names into addresses has failed, so your devices can reach nothing by name, which is the only way they know how to reach anything. The test takes a minute. If any device can be pointed at a service by its numeric address — some router admin pages and diagnostic tools allow this — or if a messaging app with established connections keeps working while every browser tab fails, names are your suspect. A quicker household test is to change nothing and simply wait ten minutes while testing a few different sites; a provider-side name service failure often heals centrally, and the pattern of “everything broken, then everything fine, with no reboot in between” is its signature in your diary.
The router-side version is more durable and more yours to fix. If the router’s name server settings were changed — by a setup wizard’s enthusiasm, an old experiment, or in the worst case by tampering — every device inherits the broken directions. Compare the DNS fields in the router’s internet settings against your household network note; if you have no note, set the router to obtain name servers automatically from the provider, restart it once, and test. Deliberately setting well-known public resolvers is the other legitimate remedy, and the security guide on this site explains the choice in full. What matters here is the diagnostic order: when the lights say the line is up and the router says it is connected, suspect names before you suspect the line, and you will skip the factory reset that this particular fault never needed. Note the episode in the post-mortem the main guide recommends, because name failures recur in clusters and the second occurrence should take you ninety seconds, not an evening.
Credentials and tags: the settings a new box forgets
Some connections authenticate, and authentication is where replacement hardware and factory resets draw their sharpest sting. On connections that use a username and password for the line itself — the PPPoE style common on many DSL and some fibre services — the router must present those credentials before the provider lets it online, and they are not your Wi-Fi password, your account password or anything else you use daily; they are a specific pair, issued at installation, often printed on a card that has been in a drawer ever since. A router reset to factory state has forgotten them, and its symptom is precisely this article’s title: lights healthy, network broadcasting, no internet, forever, until the credentials are re-entered. The same family includes services that require a numeric tag on the connection — a VLAN identifier the router must attach — without which the line stays politely dark. Both settings live in the router’s internet or WAN section, both are provider-specific, and both are un-guessable.
The prevention writes itself, and it belongs in every household’s network note: record the connection type, the line username, and any tag or special setting the provider required, on the day the service first works. Providers can re-issue credentials over the phone after identity checks, so a lost card is an inconvenience rather than a catastrophe — but it is an inconvenience scheduled for the exact evening everything is already broken. If you are reading this mid-incident with a freshly reset router and no note, call the provider before trying anything else; ask specifically for the connection credentials and settings for a third-party router, and write down everything they say, because the note you make during the bad evening is the one that prevents the next one. Households that keep their own router behind a provider gateway in bridge mode are largely immune to this whole family of faults, since the gateway holds the credentials — one more quiet argument for that architecture.
Proving an outage: status pages, neighbours and the provider’s own telemetry
The main guide’s step on ruling out the provider deserves a fuller toolkit, because “is it just me?” is the question every no-internet incident turns on, and it can now be answered from the sofa — provided you have a phone with mobile data, which is the one connection the outage cannot take. Start with the provider’s own status page and app, read with a sceptic’s eye: providers confirm large outages promptly and small ones slowly, so an all-clear there proves little in the first hour. Independent outage-reporting sites, which aggregate user reports into live graphs, are often faster for local faults; a spike of reports naming your provider in your region is strong evidence, and a flat line is weak evidence against. Your street remains the best sensor of all: a neighbour on the same provider, asked by text, settles the question for your shared infrastructure in a way no website can.
While you gather evidence, remember what not to do: do not factory-reset your router during a suspected provider outage. It cannot help — the fault is not in your house — and it converts a wait into a rebuild, because when the provider restores service you will be reconfiguring instead of reconnecting. The correct posture during a confirmed outage is a powered, untouched network and a household that knows the state of play. If the outage stretches, the provider’s estimated restoration time, however elastic, is still the best planning figure you have, and repeated power-cycling every twenty minutes adds nothing but wear to your patience. Log the start time, the provider’s reference number if you call, and the restoration time; providers’ compensation and automatic credit schemes, where they exist, run on exactly those timestamps, and the household that has them written down is the household that receives them without an argument.
Keeping a household running: fallback connections done safely
For households that cannot simply wait — the remote worker mid-deadline, the student in an exam window, the home business taking orders — a fallback connection is the difference between an outage and an inconvenience, and it is best designed before it is needed. The simplest fallback is a phone’s hotspot, used by the one or two devices that matter rather than by the whole house: laptops and phones join it directly, the television and the gadgets stay dark, and the data allowance survives the day. Know in advance whether your phone plan permits tethering and how much data it includes, because discovering a tethering block during the outage adds insult to it. The step up is a router that accepts a mobile connection as a backup — some models take a USB modem or a SIM directly, or can use a phone’s tethered connection — and switches the whole house over automatically when the main line dies, then switches back when it heals. Set up in calm weather and tested once, such a failover is invisible in use; set up for the first time during a crisis, it is another router configuration problem stacked on top of the outage.
Two cautions keep fallbacks honest. A failover connection is usually slower and always more finite than the line it replaces, so the household needs a shared understanding of the rationing: work and study first, streaming and updates paused, the console’s background downloads explicitly stopped, because one uncapped device can consume a month’s mobile data in an afternoon. And when the main line returns, confirm the house has actually switched back — most systems do this automatically, but a router left quietly running the household on mobile data for a week is an expensive silence. Note the fallback’s existence, its data budget and its test date in the network record; like the credentials in the previous section, it is a calm-weather asset that cannot be improvised in the storm. Households formalising this resilience — a second connection as a deliberate design — will find the routers that handle dual connections gracefully among Best Routers for Home Office, where continuity is a purchasing criterion rather than a hope.
After a replacement box: activation, registration and the first-day faults
A distinct species of no-internet follows the arrival of new hardware: the provider has shipped a replacement gateway, or you have bought your own modem, and the new box sits with its lights in a pattern that never quite reaches connected. The missing step is usually registration. Providers bind service to the identity of the modem or gateway on the account, and a new unit is a stranger until it is introduced — by an activation page that appears when you first try to browse, by an app flow, or by a call where you read the new unit’s identifiers from its label. If a fresh box will not come online within the time its guide promises, stop power-cycling and check for an activation step before suspecting a fault; the lights of an unregistered modem and a broken line are easy to confuse, and only one of them is fixed by a phone call. Photograph the new unit’s label for the network note while you have it in your hands — the identifiers on it are what every future support call will ask for.
The router-side first-day faults are the mirror image: a new router behind a working modem that shows no internet because the modem remembers the old router. Some modems bind themselves to the first device they meet and must be restarted — fully, with the pause the main guide describes — to accept a new partner; others need the new router to wait through the modem’s full startup before being powered itself. The order in the main guide’s power-cycle step exists for exactly this courtship: modem settled completely, then router. If a new installation defeats you after an honest hour, the provider’s activation line and the router maker’s setup support are both legitimate next steps rather than admissions of defeat — first-day faults are the most common kind in all of networking, and they are almost never a verdict on the hardware or on you.
When it is only one device saying “no internet”
Sometimes the whole house is online and a single device insists otherwise, and the household burns an evening troubleshooting a network that was never broken. One-device no-internet has its own causes, and they are almost all on the device. A stale connection is the commonest: the device holds settings from before a router change or a password rotation, and the cure is to make it forget the network entirely and join fresh, rather than toggling Wi-Fi and hoping. A device with a manually set address or name server — configured for some past experiment, office or school network — can sit on a healthy home network unable to resolve anything, and restoring its settings to automatic is a two-minute fix that no router restart will ever achieve. Date and time deserve a check on any device that refuses secure sites while loading others: a device whose clock has drifted badly, often after a flat battery, fails the security checks modern websites require and reports the failure as a connection problem.
Security software is the fourth suspect, and the most theatrical: an expired or confused firewall or VPN application on a laptop can block all traffic while announcing that the network is at fault. If one computer is dark while its siblings browse, disable the VPN first and test, then the security suite’s network components, one at a time. The household rule that emerges is worth stating in the network note: a fault reported by exactly one device is investigated on that device, and the router is touched only when a second device agrees with the first. More evenings are lost to rebooting healthy routers on behalf of one confused laptop than to any fault in this guide.
After the bang: checking for storm and surge damage safely
The main guide covers outages after storms; the darker variant is the storm that leaves the power on and the internet dead, with hardware damage on the table. Work safely and in order. If any equipment smells of burning, shows scorch marks, or behaves erratically, switch it off at the wall and leave it off — damaged power supplies are a fire question, not a networking question, and no diagnosis is worth continuing on a box you do not trust. Otherwise, triage by substitution with what the house already owns: the modem’s lights tell you whether the line side survived; a router that will not power on with its own adapter but wakes on a matched spare has lost an adapter, the storm’s commonest victim; a router that powers on but whose Ethernet ports no longer see the modem may have taken the surge through the cable path, which is the route lightning prefers into home networks.
Providers replace their own storm-killed gateways as a matter of routine, and the call goes better with the same evidence discipline as any other in this guide: the time of the storm, the symptoms in order, and what substitution tests you have already run. Your own router, if it is the casualty, is a purchase rather than a call — and the replacements in Best Routers for Home Use are the sensible shortlist for a household buying under time pressure. Let the incident argue for one upgrade regardless of what it killed: a surge protector between the wall and the network shelf, replaced on the schedule its maker recommends, is cheap insurance for the equipment the whole house leans on, and the day you install it is a good day to photograph every adapter’s label for the network note, so the next substitution test takes minutes.
Storms, credentials, name servers, one confused laptop — the variety inside this single symptom is exactly why the ordered approach matters. Work the splits in sequence, write down what each one showed, and the fault runs out of places to hide: every no-internet incident ends with a cause, and most end within the hour, with the household’s record of it making the next one shorter.
Keep that record somewhere the whole household can find it. The next outage will not ask whether you are home; it will ask whether anyone in the house can read the lights, find the note and start the sequence — and that, more than any single fix in this article, is what a well-kept network actually is.
Write it once, on a good day, and every bad day afterwards gets measurably, reliably shorter.
Frequently asked questions
Why does my router say ‘connected, no internet’?
That message means the device joined your Wi-Fi fine, but the router has no working path onward — usually the modem has lost sync, the modem-router link is broken, or the provider is down. The fault is upstream of your Wi-Fi, which is why the lights and the wired test in this guide find it.
Will restarting really fix no internet?
Very often, yes — when done in the right order with a proper wait. The modem must fully re-establish its provider connection before the router asks it for one; restarting only the router, or powering both at once, leaves the mismatch in place.
How long should I wait for the modem to come back?
Up to five minutes for its online light to go solid is normal on cable and DSL. If it is still blinking after ten minutes and a second power cycle, treat it as a line or modem fault and contact the provider.
My internet works wired but not on Wi-Fi. What now?
The connection is fine; the router’s wireless side is not. Check that both bands are broadcasting, forget and rejoin the network on one device as a test, and update the firmware. A single vanished band after an update is a known pattern — the update guide covers recovery.
Should I buy a new router if this keeps happening?
Only after the pattern convicts the router: wired bypass works, settings are right, firmware is current, and drops persist. If the modem’s lights are the ones failing, a new router changes nothing — that is the provider’s line to fix.
Everything says “connected” but nothing loads. Is that different from no internet?
Yes, and the difference steers the fix. “Connected, no internet” on your devices means they can see the router but the router cannot vouch for the wider connection — work through this guide’s main steps, starting with the lights. A connection where pages begin loading and stall, or some services work while others fail, points instead at the impostor faults: name resolution, a saturated upload, or an application-specific block. Test a few different services, note exactly which fail, and let that pattern choose your section rather than restarting equipment on instinct.
How long should I wait before calling the provider?
After you have done the ordered basics — lights read, one correct power-cycle, cables checked, a wired test attempted — and gathered the outage evidence from status pages and neighbours, calling is not impatience; it is the next step. That sequence takes about fifteen minutes, and arriving at the call with it done moves you past the script’s first page immediately. Call sooner if there is any sign of physical damage, a burning smell, a line down outside, or if the household includes medical or security equipment that depends on the connection — and say so at the start of the call, because providers prioritise accordingly.
The internet came back on its own. Should I still do anything?
Yes: the five-minute post-mortem from the main guide, while the details are fresh. Note the start and end times, the lights’ behaviour, anything you changed, and whether the pattern matches any previous entry. Self-healing incidents are usually provider-side blips, and a single one needs no further action. A diary showing the same self-healing outage every few weeks is a different matter — that is an intermittent line fault introducing itself politely, and the diary is what turns your eventual report from a shrug into a work order.
Can a faulty Ethernet cable really cause a total no-internet?
Completely — the single cable between the modem and the router is the one cable whose failure blacks out the whole house while every light on the Wi-Fi side looks cheerful. Cable faults hide in pinched sections behind furniture, in plugs with broken clips that creep halfway out, and in the flexible patch lead that has been bent once too often. Swapping that one cable for a known-good spare is the cheapest test in this entire guide, which is why the main steps place cable checks before settings expeditions: two minutes, no menus, and a surprising share of total outages end there.
Related Guides
- How to Restart a Router — the power-cycle procedure this guide builds on
- Modem vs Router — the two-box model behind the isolation test
- How to Fix a Router that Keeps Disconnecting — the sibling fault: connected, but dropping
- How to Set Up a Router — what the last-resort reset obliges you to redo


