Post · bonfire.cafe

Can’t connect to public WiFi via Linux

I am using Fedora 42, and I am unable to access the captive portal of inOui TGV. I can connect to the WiFi with no problems, but I can’t get access to the Internet.

I have tried everything:

Connecting to 192.168.1.1, to http://neverssl.com, to http://wifi.sncf via browser
Running sudo dhclient -r and rebooting the device
Running systemctl restart NetworkManager
Also other stuff I tried in the past with other public WiFis. I am sure there is something wrong with my specific device configuration, it’s not this network.

Nothing works 😩

Can somebody who understands this stuff help me?

#help #mastoAiuto #Linux #LinuxNetwork #Network #CaptivePortal #WiFi #PublicWiFi #HTTP #HTTPS #troubleshoot #LinuxHelp #Fedora #Fedora42

Haralds Jakovels

@hjakovel@toot.lv replied · 3 weeks ago

@tommi in their webpage, they mention address https://wifi.sncf, you say you were connecting to http://wifi.sncf, is this correct?

john r red-horse

@jrredho@mastodon.world replied · 4 weeks ago

@tommi

I've seen posts in the https://discussion.fedoraproject.org discussion forum where wifi problems came down to using wpa_supplicant vs iwd as the interface software for it.

I'd strongly suggest taking a look there, and perhaps posting a question yourself if you have trouble finding a suitable answer to this.

Good luck!

medium rare bird

@migratory@jorts.horse replied · 4 weeks ago

@tommi my recent experience with captive portals and linux has been that many captive portals' log-in pages are served with broken DNS that can only be resolved using the DNS server provided by DHCP from the router hosting the captive portal. if you instead use a "sans" DNS server (1.1.1.1 or 8.8.8.8 or 4.2.2.1 or whatever browsers reach via DNS-over-HTTP) you won't be able to reach the portal login page. you might need to add the captive portal's one to /etc/resolv.conf

Philipp Memmel

@PhilippMemmel@bildung.social replied · 4 weeks ago

@migratory @tommi just throwing in another thing here, even though I don't think it might be related as ip and routing looks good: I had similar issues because of conflicting subnets of virtual docker devices. Do you happen to use docker on this machine?

Jonathan

@jonathan@charade.social replied · 4 weeks ago

@tommi try attempting to visit 1.1.1.1 in a web browser

N0tSure

@N0tSure@mastodon.social replied · 4 weeks ago

@tommi It happens to me once in a while. I typically use my phone to connect and bluetooth to the computer. I think it happens to routers that are never reset.

Jupiter

@avoca@gladtech.social replied · 4 weeks ago

@tommi

I have the same problem with public WiFi at my local shopping centre.

I have concluded the centre's system won't run with Linux for tracking/security reasons.

I just connect via my iOS phone and hotspot my Linux laptop through that, works flawlessly.

NegativePrimes

@negativeprimes@urusai.social replied · 4 weeks ago

@tommi I'm no expert, but the other day I tried connecting to a public network and nothing worked... until I rebooted. Then everything was magically fine. FWIW.

Tommi 🤯

@tommi@pan.rent replied · 4 weeks ago

@negativeprimes Did you do something before rebooting?

NegativePrimes

@negativeprimes@urusai.social replied · 4 weeks ago

@tommi I had updated the system, and apparently something borked until the reboot. I didn't try much else before that, just toggling VPNs and switching to different networks to see if that would make a difference (it did not). Also ran "ip link" to view the status of my network interfaces.

Earth Notes

@EarthOrgUK@mastodon.energy replied · 4 weeks ago

@tommi What I generally find breaks public WiFi for me on my Mac is having set/forced DNS to public resolvers such as 1.1.1.1 and 8.8.8.8 which breaks whatever nonsense goe on in the start-up dance with such a portal. Once I remove the forced DNS resolvers everything is good again, usually.

FandaSin

@FandaSin@social.linux.pizza replied · 4 weeks ago

@tommi

What you get for command (it should print out all IPs and devices)
"ip -a"
and
"ip route"
(routing of your network)?

Might be also problem with DNS, but lets start with basics. (connection and routing)

Tommi 🤯

@tommi@pan.rent replied · 4 weeks ago

Thanks for replying, @FandaSin and all!

Here are the outputs of several info commands:

ip route:

default via 10.17.0.2 dev wlp1s0 proto dhcp src 10.17.4.41 metric 600
10.17.0.0/21 dev wlp1s0 proto kernel scope link src 10.17.4.41 metric 600

ip a:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute
       valid_lft forever preferred_lft forever
2: wlp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
    link/ether 86:67:7f:b5:b8:e1 brd ff:ff:ff:ff:ff:ff permaddr 14:ac:60:29:7d:17
    altname wlx14ac60297d17
    inet 10.17.4.41/21 brd 10.17.7.255 scope global dynamic noprefixroute wlp1s0
       valid_lft 101sec preferred_lft 101sec
    inet6 fe80::747e:df9:ee7e:71c0/64 scope link noprefixroute
       valid_lft forever preferred_lft forever

ifconfig:

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 1000  (Local Loopback)
        RX packets 1264  bytes 95511 (93.2 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 1264  bytes 95511 (93.2 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

wlp1s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 10.17.4.41  netmask 255.255.248.0  broadcast 10.17.7.255
        inet6 fe80::747e:df9:ee7e:71c0  prefixlen 64  scopeid 0x20<link>
        ether 86:67:7f:b5:b8:e1  txqueuelen 1000  (Ethernet)
        RX packets 12653  bytes 744838 (727.3 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 7864  bytes 665881 (650.2 KiB)
        TX errors 0  dropped 25 overruns 0  carrier 0  collisions 0

resolve.conf:

nameserver 127.0.0.53
options edns0 trust-ad
search .

FandaSin

@FandaSin@social.linux.pizza replied · 4 weeks ago

@tommi

Sorry I was just on lunch...

This looks good.
Then I would try to "ping 10.17.0.2"
But I suppose it should work without problems.

Maybe try "traceroute 8.8.8.8" (or 4.4.4.4, or some other IP outside your network)

I would say it might be problem with some strange DNS setting?

mkj

@mkj@social.mkj.earth replied · 4 weeks ago

@tommi `ip -a` might have been intended to be `ip a` (or `ip address` if you aren't trying to save a few characters).

@FandaSin

mkj

@mkj@social.mkj.earth replied · 4 weeks ago

@FandaSin beat me to it.

Start by checking that the IP addresses and ideally also routing table is reasonable. Try `ping -n` the default gateway by IP address. (`-n` turns off reverse DNS lookups.)

Check /etc/resolv.conf to make sure it has correct nameservers. Try `ping -n` those by IP address, see if they respond.

`nmcli general status` and `nmcli networking connectivity check` and `nmcli connection show` might also be helpful.

@tommi

Thibaultmol 🌈

@thibaultmol@en.osm.town replied · 4 weeks ago

@tommi Gonna following this, cause I also recently had a problem. I tried going to neverssl.com to trigger an http redirect and while it redirect, it doesn't load the captive portal site of the hospital I was at

Steve's Place

@steter@mastodon.stevesworld.co replied · 4 weeks ago

@thibaultmol @tommi I don't use Fedora, so I might be useless for this, but it sounds like you might be connecting to an open wifi, not your router. If that's the case, you'll need to get the password and wifi name to use it. It does your routing, and should provide you with nameservers for DNS.

Try this:

ping 1.1.1.1

If you can't ping them, you aren't connected to the internet. If you can ping them, you are connected, but have a DNS problem (which would become apparent if you tried to ping yahoo.com and got nothing back). Perhaps your router doesn't provide nameservers. You could set 1.1.1.1 and 1.0.0.1 as your nameservers. They are public, and always available.

As I said, I don't use Fedora (it's been years). Just a guess.

Thibaultmol 🌈

@thibaultmol@en.osm.town replied · 4 weeks ago

@steter @tommi i tried that, i couldn't get a ping back. (Tommi's situation might be different tbf. Sorry for piggybacking on this)

In my case it was public wifi in a hospital. No password on the wifi but pin code on the captive portal.

In tommi's case it's also with a captive portal iirc

bonfire.cafe

A space for Bonfire maintainers and contributors to communicate

bonfire.cafe: About · Code of conduct · Privacy · Users · Instances

Bonfire social · 1.0.0-rc.2.11 no JS en

Automatic federation enabled