[SOLVED] Seemingly random connection errors

idontgetit · 2025-08-25 19:32:57

Hi,

Hopefully I'm not posting a duplicate. I tried searching the forum and GitLab but was unable to find anything. If this has already been solved I apologize.

In any case, lately I've been experiencing some weird network issues (random connection drops or refusals). For example:

1. Firefox says it's "unable to connect" (NS_ERROR_CONNECTION_REFUSED), requiring several page refreshes after which the browser finally manages to load the page
2. Packet loss in ping (5 % and up)
3. MTR errors (mtr: Unexpected mtr-packet error)

All of this seems very random. The connection might work fine for an hour, then experience some errors for a couple of minutes, and then resume normal operation.

The thing is, all of this seems somehow Arch-related. I've got two Arch installations (ThinkPad laptop and a custom-built desktop) experiencing these issues. However, no other computer in the household seems to be affected. Three Windows PCs work just fine. Even virtualized Linux Mint guest on an Arch host works fine (in other words: while Arch host experiences constant mtr errors, mtr in Linux Mint guest works without a hitch).

Things that I've tried:
1. Default kernel (6.16.3) and LTS (6.12.43)
2. NetworkManager and netctl
3. Dynamic and static addresses
4. Different DNS servers
5. Router reset
6. Different router
7. Cable and wireless

I've got very basic networking setup. Basically just NetworkManager on automatic and ufw with SSH allowed. Also, I'm not able to find anything related in journalctl -xe or dmesg (mostly just UFW BLOCK noise but that I've had before, too).

Originally I was thinking this could be a hardware issue but it seems at least one other person is experiencing something similar: https://www.reddit.com/r/arch/comments/ … etworking/

Any ideas what to try next?

Last edited by idontgetit (2025-09-08 21:32:51)

Bartoszer · 2025-08-25 20:52:15

Hello, I'm also getting similar weird network errors.

Kernel 6.16.3-arch1-1
Using netctl for network.

I noticed one thing. One of the games I play recently throws shitloads of errors
[t=00:29:58.195093][f=0034244] Warning: Network error 13: Permission denied

Looks like some sort of permission issue after upgrading the system.

No idea how to debug it though

gugah · 2025-08-25 20:58:32

NetworkManager w/ kernel 6.16.3 here, same problems. For example:

$ ping github.com
PING github.com (140.82.114.3) 56(84) bytes of data.
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=1 ttl=50 time=36.2 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=2 ttl=50 time=32.0 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=3 ttl=50 time=33.0 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=4 ttl=50 time=31.4 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=5 ttl=50 time=37.6 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=6 ttl=50 time=31.9 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=7 ttl=50 time=30.8 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=8 ttl=50 time=34.9 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=9 ttl=50 time=37.3 ms
64 bytes from lb-140-82-114-3-iad.github.com (140.82.114.3): icmp_seq=10 ttl=50 time=32.7 ms
^C
--- github.com ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9011ms
rtt min/avg/max/mdev = 30.812/33.782/37.620/2.402 ms

When trying again, seemingly at random:

$ ping github.com
ping: Do you want to ping broadcast? Then -b. If not, check your local firewall rules

Last edited by gugah (2025-08-25 21:08:40)

Bartoszer · 2025-08-25 21:08:27

https://bbs.archlinux.org/viewtopic.php?id=307778

Seems related

idontgetit · 2025-08-26 14:45:07

Hmm. I wonder if there are different issues at play here.

In any case, I did a simple curl loop trying out different hosts, 1 / sec.

Here's what the current kernel 6.16.3 gives me:

OK 1.1.1.1
OK 8.8.8.8
OK google.com
curl: (35) Recv failure: Connection reset by peer
OK microsoft.com
OK cloudflare.com
curl: (7) Failed to connect to redhat.com port 443 after 2 ms: Could not connect to server
curl: (7) Failed to connect to ubuntu.com port 443 after 9 ms: Could not connect to server
curl: (7) Failed to connect to apple.com port 443 after 30 ms: Could not connect to server
OK oracle.com
OK wikipedia.org
curl: (7) Failed to connect to facebook.com port 443 after 8 ms: Could not connect to server
OK openai.com
OK amazon.com

Then I booted an installation USB I made earlier this month (kernel 6.15.8) and executed the same script. Only OKs, not a single "failed to connect".

After bit of experimenting I downgraded the kernel of the actual system to the same one as on the installation USB. Result: curl works fine, no failed connections.

But what baffles me is that Linux LTS (6.12.43) has the same errors as 6.16.3 (that is, curl has several failures with LTS, too).

I'm not sure about mtr and other tests yet, they're a bit more difficult to catch, anyway.

seth · 2025-08-26 14:51:57

Please post the output of

find /etc/systemd -type l -exec test -f {} \; -print | awk -F'/' '{ printf ("%-40s | %s\n", $(NF-0), $(NF-1)) }' | sort -f

Please post your complete system journal for the boot:

sudo journalctl -b | curl -F 'file=@-' 0x0.st

after some connection failures.

ponsilio · 2025-08-26 15:18:58

I am experiencing the exact same problem in 2 laptops, in 2 different locations/networks.

Here's the journalctl near one of the errors (trying to load this very same page nonetheless):
http://0x0.st/Koq5.txt

And here's the output of the first command:

avahi-daemon.service                     | multi-user.target.wants
avahi-daemon.socket                      | sockets.target.wants
bluetooth.service                        | bluetooth.target.wants
cups.path                                | multi-user.target.wants
cups.service                             | multi-user.target.wants
cups.service                             | printer.target.wants
cups.socket                              | sockets.target.wants
dbus-org.bluez.service                   | system
dbus-org.freedesktop.Avahi.service       | system
dbus-org.freedesktop.nm-dispatcher.service | system
dbus-org.freedesktop.timesync1.service   | system
default.target                           | system
display-manager.service                  | system
docker.socket                            | sockets.target.wants
fstrim.timer                             | timers.target.wants
getty@tty1.service                       | getty.target.wants
libvirtd-admin.socket                    | sockets.target.wants
libvirtd-ro.socket                       | sockets.target.wants
libvirtd.service                         | multi-user.target.wants
libvirtd.socket                          | sockets.target.wants
NetworkManager.service                   | multi-user.target.wants
NetworkManager-wait-online.service       | network-online.target.wants
nmb.service                              | multi-user.target.wants
p11-kit-server.socket                    | sockets.target.wants
paccache.timer                           | timers.target.wants
pamac-cleancache.timer                   | timers.target.wants
pipewire-pulse.socket                    | sockets.target.wants
pipewire-session-manager.service         | user
pipewire.socket                          | sockets.target.wants
power-profiles-daemon.service            | graphical.target.wants
remote-fs.target                         | multi-user.target.wants
smb.service                              | multi-user.target.wants
systemd-timesyncd.service                | sysinit.target.wants
systemd-userdbd.socket                   | sockets.target.wants
virtlockd-admin.socket                   | sockets.target.wants
virtlockd.socket                         | sockets.target.wants
virtlogd-admin.socket                    | sockets.target.wants
virtlogd.socket                          | sockets.target.wants
wakeup-triggers.service                  | multi-user.target.wants
wireplumber.service                      | pipewire.service.wants
xdg-user-dirs-update.service             | default.target.wants

Last edited by ponsilio (2025-08-26 17:12:38)

ponsilio · 2025-08-26 15:27:42

By the way, I am actually in EndevourOS (not "pure" Arch Linux) and located in Barcelona.

I mention the location because there have been notorious network related problems due to blocking imposed by judges due to requests from football broadcasters. However this seems unrelated because:
1-It appears to be random whereas deliberate blocking can be replicated.
2-Other operating systems seem unaffected AFAIK (namely Windows/Android/iOS).

seth · 2025-08-26 15:37:43

Please use [code][/code] tags. Edit your post in this regard.

Whether it's the same we'll see, but you're initially connected for 5s via the iwlifi nic and then switch to some usb dongle (or tethering device) and the only disconnects seem from suspend cycles.
1. What is actually the affected NIC
2.

namely Windows

3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.
3. The docker devices might get in the way, check "ip r get 8.8.8.8" to see where you're actually routing when this happens.

For the usb dongle, see https://wiki.archlinux.org/title/Power_ … utosuspend

TheCaRp · 2025-08-26 16:12:09

I had the same issue. Firefox not connecting. Connection errors in games. Etc. I downgraded to Kernel 6.16.1-arch1-1 and everything is back to normal.

Related: https://bbs.archlinux.org/viewtopic.php … 2#p2258072

Last edited by TheCaRp (2025-08-26 16:18:55)

ponsilio · 2025-08-26 17:21:29

seth wrote:

Please use [code][/code] tags. Edit your post in this regard.
Whether it's the same we'll see, but you're initially connected for 5s via the iwlifi nic and then switch to some usb dongle (or tethering device) and the only disconnects seem from suspend cycles.
1. What is actually the affected NIC
2.
namely Windows
3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.
3. The docker devices might get in the way, check "ip r get 8.8.8.8" to see where you're actually routing when this happens.
For the usb dongle, see https://wiki.archlinux.org/title/Power_ … utosuspend

I added the code tags as instructed.
The NIC affected is a USB ethernet dongle (enp0s20f0u1u4).
I don't have Windows in this exact computer. When I wrote that it does not happen in Windows I meant in other systems. Basically I wanted to pinpoint the fact that it does not seem to be related to the network per se (router, ISP).

ip r get 8.8.8.8
8.8.8.8 via 192.168.1.1 dev enp0s20f0u1u4 src 192.168.1.53 uid 1000

It seems to be using the proper routing.
It might be related to USB autosuspend. Long time ago I played a bit with it but it has been working fine for many months without any change from my part.
BTW: When the computer boots it autoconnects to a wifi network that I manually disconnect to ensure the wired connection is used. That's why you see the disconnecting in the logs short after a boot up.

Other info: some sites seems to be more prone to cause the error. One is precisely bbs.archlinux.org and other is the Google Calendar. Might be a coincidence, though.

I will try to switch to wifi (despite being a bit slower) and see if the problem persists.

ponsilio · 2025-08-26 18:26:03

I can confirm that using the internal wifi card exhibits the same behavior.
Does not seem related to USB autosuspend because the internal wifi card is connected through pci.

Some more context:
* When a page is not loading properly, just trying to load it again seems to fix the problem immediately.
* I tried different DNS servers and the behavior does not change. I did that because some Python script I use regularly indicated that it could not resolve the name with the message "Temporary failure in name resolution". Also this hints the fact that the problem is system wide. By that I mean that it is not caused by a specific browser.

seth · 2025-08-26 19:08:19

Do you get this w/ the 6.16.1 kernel?

ponsilio · 2025-08-26 19:52:00

seth wrote:

Do you get this w/ the 6.16.1 kernel?

So far I tried the latest kernel (6.16.3) and the LTS one 6.12.43-1-lts. Happens in both kernels.
I don't know how to try a specific kernel but I will look it up and test it.

However is strange that even LTS shows this behavior. This suggests that the main problem might not be in the kernel.

gugah · 2025-08-26 20:00:11

just downgraded to 6.16.1. No issues so far.

$ ./curl_test.sh  
OK 1.1.1.1
OK 8.8.8.8
OK google.com
OK microsoft.com
OK cloudflare.com
OK redhat.com
OK ubuntu.com
OK apple.com
OK oracle.com
OK wikipedia.org
OK facebook.com
OK openai.com
OK amazon.com

Just like @ponsillo mentioned, seems to happen with some sites more often than others and it's system-wide (as in browsers, pacman mirrors when running pacman -Syu, ping, curl.

seth · 2025-08-26 20:05:06

However is strange that even LTS shows this behavior.

https://lore.kernel.org/regressions/202 … ecast.net/
https://cdn.kernel.org/pub/linux/kernel … Log-6.16.2
https://cdn.kernel.org/pub/linux/kernel … og-6.12.43
Search the latter two for "9e30ecf23b1b8f091f7d08b27968dea83aae7908" …

ponsilio · 2025-08-26 20:15:15

I just downgraded the kernel to 6.16.1, rebooted (ensured I'm actually running 6.16.1) and... the first time I tried to access this very same forum thread through a bookmark... bam, the error again.
Just pressing the bookmark again loads the page just fine.

I don't think it's the kernel is the culprit, guys. If that where the case the LTS one should work just fine. Unless some other change has make a kernel bug surface.
Also it happens with different laptop models and both with USB ethernet and wifi cards (different drivers).

I noticed the problem started like 3 or 4 days ago for me. So I listed the packages updated on the last 7 days:

[2025-08-19T20:26:01+0200] [ALPM] upgraded ell (0.78-1 -> 0.79-1)
[2025-08-19T20:26:01+0200] [ALPM] upgraded eos-translations (25.6-1 -> 25.8-1)
[2025-08-19T20:26:01+0200] [ALPM] upgraded hdparm (9.65-2 -> 9.65-3)
[2025-08-19T20:26:01+0200] [ALPM] upgraded hwdata (0.397-1 -> 0.398-1)
[2025-08-19T20:26:01+0200] [ALPM] upgraded libgcrypt (1.11.1-1 -> 1.11.2-1)
[2025-08-19T20:26:01+0200] [ALPM] upgraded libinput (1.28.1-1 -> 1.29.0-1)
[2025-08-19T20:26:01+0200] [ALPM] upgraded kfilemetadata (6.17.0-1 -> 6.17.0-2)
[2025-08-19T20:26:02+0200] [ALPM] upgraded python-distlib (0.3.9-2 -> 0.4.0-1)
[2025-08-19T20:26:02+0200] [ALPM] upgraded python-fastjsonschema (2.21.1-1 -> 2.21.2-1)
[2025-08-19T20:26:02+0200] [ALPM] upgraded python-orjson (3.11.1-1 -> 3.11.2-1)
[2025-08-19T20:26:02+0200] [ALPM] upgraded python-requests (2.32.4-1 -> 2.32.5-1)
[2025-08-19T20:26:02+0200] [ALPM] upgraded python-virtualenv (20.28.0-1 -> 20.34.0-1)
[2025-08-19T20:26:02+0200] [ALPM] upgraded rpcbind (1.2.7-2 -> 1.2.8-1)
[2025-08-19T20:26:29+0200] [ALPM] upgraded visual-studio-code-bin (1.103.0-1 -> 1.103.1-1)
[2025-08-19T20:26:29+0200] [ALPM] upgraded google-chrome (139.0.7258.66-1 -> 139.0.7258.127-1)
[2025-08-20T19:48:51+0200] [ALPM] upgraded libavc1394 (0.5.4-6 -> 0.5.4-7)
[2025-08-20T19:48:51+0200] [ALPM] upgraded firefox (141.0.3-1 -> 142.0-1)
[2025-08-20T19:48:51+0200] [ALPM] upgraded firefox-i18n-ca (141.0.3-1 -> 142.0-1)
[2025-08-20T19:48:51+0200] [ALPM] upgraded libatasmart (0.19-6 -> 0.19-7)
[2025-08-20T19:48:51+0200] [ALPM] upgraded meson (1.8.3-1 -> 1.8.4-1)
[2025-08-20T19:48:52+0200] [ALPM] upgraded sof-firmware (2025.05-1 -> 2025.05.1-1)
[2025-08-20T19:48:52+0200] [ALPM] upgraded spandsp (0.0.6-5 -> 0.0.6-6)
[2025-08-20T19:48:52+0200] [ALPM] upgraded startup-notification (0.12-8 -> 0.12-9)
[2025-08-23T08:57:18+0200] [ALPM] upgraded gdbm (1.25-1 -> 1.26-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded bind (9.20.11-1 -> 9.20.12-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded cryptsetup (2.8.0-1 -> 2.8.1-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded edk2-aarch64 (202411-1 -> 202505-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded edk2-arm (202411-1 -> 202505-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded edk2-ovmf (202411-1 -> 202505-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded harfbuzz (11.4.1-1 -> 11.4.2-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded harfbuzz-icu (11.4.1-1 -> 11.4.2-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded ldb (2:4.22.3-1 -> 2:4.22.4-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded lib32-harfbuzz (11.4.1-1 -> 11.4.2-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded mesa (1:25.1.7-1 -> 1:25.2.1-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded lib32-mesa (1:25.1.7-1 -> 1:25.2.1-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded lib32-sqlite (3.50.2-1 -> 3.50.4-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded vulkan-intel (1:25.1.7-1 -> 1:25.2.1-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded lib32-vulkan-intel (1:25.1.7-1 -> 1:25.2.1-1)
[2025-08-23T08:57:19+0200] [ALPM] upgraded libwbclient (2:4.22.3-1 -> 2:4.22.4-1)
[2025-08-23T08:57:20+0200] [ALPM] upgraded linux (6.16.1.arch1-1 -> 6.16.2.arch1-1)
[2025-08-23T08:57:21+0200] [ALPM] upgraded linux-headers (6.16.1.arch1-1 -> 6.16.2.arch1-1)
[2025-08-23T08:57:21+0200] [ALPM] upgraded node-gyp (11.3.0-1 -> 11.4.1-1)
[2025-08-23T08:57:21+0200] [ALPM] upgraded python-coverage (7.10.3-1 -> 7.10.4-1)
[2025-08-23T08:57:21+0200] [ALPM] upgraded python-lxml (6.0.0-2 -> 6.0.1-1)
[2025-08-23T08:57:22+0200] [ALPM] upgraded qt6-declarative (6.9.1-2 -> 6.9.1-3)
[2025-08-23T08:57:22+0200] [ALPM] upgraded smbclient (2:4.22.3-1 -> 2:4.22.4-1)
[2025-08-23T08:57:22+0200] [ALPM] upgraded samba (2:4.22.3-1 -> 2:4.22.4-1)
[2025-08-23T08:57:22+0200] [ALPM] upgraded thin-provisioning-tools (1.2.0-1 -> 1.2.1-1)
[2025-08-23T08:57:22+0200] [ALPM] upgraded vulkan-mesa-layers (1:25.1.7-1 -> 1:25.2.1-1)
[2025-08-23T08:58:35+0200] [ALPM] upgraded dropbox (229.4.5197-1 -> 230.4.8797-1)
[2025-08-23T08:58:35+0200] [ALPM] upgraded google-chrome (139.0.7258.127-1 -> 139.0.7258.138-1)
[2025-08-24T16:32:45+0200] [ALPM] upgraded cmake (4.0.3-1 -> 4.1.0-1)
[2025-08-24T16:32:45+0200] [ALPM] upgraded hplip (1:3.25.2-1 -> 1:3.25.6-1)
[2025-08-24T16:32:45+0200] [ALPM] upgraded mesa (1:25.2.1-1 -> 1:25.2.1-2)
[2025-08-24T16:32:46+0200] [ALPM] upgraded libreoffice-fresh (25.2.5-2 -> 25.8.0-1)
[2025-08-24T16:32:47+0200] [ALPM] upgraded linux (6.16.2.arch1-1 -> 6.16.3.arch1-1)
[2025-08-24T16:32:48+0200] [ALPM] upgraded linux-headers (6.16.2.arch1-1 -> 6.16.3.arch1-1)
[2025-08-24T16:32:48+0200] [ALPM] upgraded nano (8.5-1 -> 8.6-1)
[2025-08-24T16:32:48+0200] [ALPM] upgraded nfsidmap (2.8.3-2 -> 2.8.3-3)
[2025-08-24T16:32:48+0200] [ALPM] upgraded nfs-utils (2.8.3-2 -> 2.8.3-3)
[2025-08-24T16:32:49+0200] [ALPM] upgraded python-botocore (1.39.9-1 -> 1.40.16-1)
[2025-08-24T16:32:49+0200] [ALPM] upgraded python-coverage (7.10.4-1 -> 7.10.5-1)
[2025-08-24T16:32:49+0200] [ALPM] upgraded python-cryptography (45.0.5-1 -> 45.0.6-1)
[2025-08-24T16:32:49+0200] [ALPM] upgraded python-filelock (3.18.0-1 -> 3.19.1-1)
[2025-08-24T16:32:49+0200] [ALPM] upgraded qca-qt5 (2.3.10-2 -> 2.3.10-3)
[2025-08-24T16:32:49+0200] [ALPM] upgraded qca-qt6 (2.3.10-2 -> 2.3.10-3)
[2025-08-24T16:32:49+0200] [ALPM] upgraded qt6-webengine (6.9.1-2 -> 6.9.1-3)
[2025-08-24T16:32:49+0200] [ALPM] upgraded vulkan-intel (1:25.2.1-1 -> 1:25.2.1-2)
[2025-08-24T16:32:49+0200] [ALPM] upgraded vulkan-mesa-layers (1:25.2.1-1 -> 1:25.2.1-2)
[2025-08-24T16:32:51+0200] [ALPM] upgraded wine (10.12-1 -> 10.13-1)
[2025-08-24T16:32:51+0200] [ALPM] upgraded yad (14.1-2.1 -> 14.1_eos-2)
[2025-08-24T18:47:52+0200] [ALPM] upgraded libxmlb (0.3.22-1 -> 0.3.23-1)
[2025-08-24T18:47:52+0200] [ALPM] upgraded harfbuzz (11.4.2-1 -> 11.4.3-1)
[2025-08-24T18:47:52+0200] [ALPM] upgraded appstream (1.0.5-2 -> 1.0.6-1)
[2025-08-24T18:47:52+0200] [ALPM] upgraded appstream-qt (1.0.5-2 -> 1.0.6-1)
[2025-08-24T18:47:53+0200] [ALPM] upgraded aws-cli (1.41.9-1 -> 1.42.16-1)
[2025-08-24T18:47:53+0200] [ALPM] upgraded cython (3.1.2-1 -> 3.1.3-1)
[2025-08-24T18:47:53+0200] [ALPM] upgraded harfbuzz-icu (11.4.2-1 -> 11.4.3-1)
[2025-08-24T18:47:53+0200] [ALPM] upgraded intel-gmmlib (22.7.2-1 -> 22.8.1-1)
[2025-08-24T18:47:53+0200] [ALPM] upgraded intel-media-driver (25.2.5-1 -> 25.2.6-1)
[2025-08-24T18:47:54+0200] [ALPM] upgraded kwin (6.4.4-1 -> 6.4.4-4)
[2025-08-24T18:47:54+0200] [ALPM] upgraded lib32-harfbuzz (11.4.2-1 -> 11.4.3-1)
[2025-08-24T18:47:54+0200] [ALPM] upgraded libodfgen (0.1.8-4 -> 0.1.8-5)
[2025-08-24T18:47:54+0200] [ALPM] upgraded liborcus (0.20.1-1 -> 0.20.2-1)
[2025-08-24T18:47:54+0200] [ALPM] upgraded libqalculate (5.6.0-1 -> 5.7.0-1)
[2025-08-24T18:47:54+0200] [ALPM] upgraded libraqm (0.10.2-1 -> 0.10.3-1)
[2025-08-24T18:47:54+0200] [ALPM] upgraded libreoffice-fresh-ca (25.2.5-1 -> 25.8.0-1)
[2025-08-24T18:48:12+0200] [ALPM] upgraded visual-studio-code-bin (1.103.1-1 -> 1.103.2-1)

Both yesterday and today I updated other packages but I omitted them because the culprit must be a previous package.

Looking at the list, any suspect you might think of? Perhaps the "bind" package? (I'm no expert). Perhaps I can try to downgrade suspicious packages.

seth · 2025-08-26 20:19:17

I tried to access this very same forum thread

Doesn't count, that's because of the DDOs, keep an eye on the overall situation and other web pages.

gugah · 2025-08-26 20:57:16

@ponsillo I'm using this simple script which doesn't curl any archlinux.org website (status.archlinux.org show some still down):

#!/bin/bash

hosts=(
    "1.1.1.1"
    "8.8.8.8"
    "google.com"
    "microsoft.com"
    "github.com"
    "cloudflare.com"
    "redhat.com"
    "ubuntu.com"
    "apple.com"
    "oracle.com"
    "wikipedia.org"
    "facebook.com"
    "openai.com"
    "amazon.com"
)

for host in "${hosts[@]}"; do
    # Try to curl the host with a 5-second timeout, silent output, and show error
    output=$(curl -sS --max-time 5 "https://$host" -o /dev/null)
    status=$?
    if [ $status -eq 0 ]; then
        echo "OK $host"
    else
        # Run curl again to capture and print the error message
        curl -v --max-time 5 "https://$host" -o /dev/null 2>&1 | grep -E "curl:|Failed to connect|Recv failure"
    fi
    sleep 1
done

idontgetit · 2025-08-27 10:30:35

Oh man, this is really elusive at times. Was gone for a while, now started happening again.

> uname -a
Linux foo 6.16.3-arch1-1 #1 SMP PREEMPT_DYNAMIC Sat, 23 Aug 2025 15:32:49 +0000 x86_64 GNU/Linux

I have not experienced this with 6.15.8 that I copied from the installation USB. I did test it couple of times back and forth so that the current and LTS always had failures while 6.15.8 did not. But it could be just a wonky coincidence.

seth wrote:

Please post the output of

find /etc/systemd -type l -exec test -f {} \; -print | awk -F'/' '{ printf ("%-40s | %s\n", $(NF-0), $(NF-1)) }' | sort -f

Here goes:

cpupower.service                         | multi-user.target.wants
cups.path                                | multi-user.target.wants
cups.service                             | multi-user.target.wants
cups.service                             | printer.target.wants
cups.socket                              | sockets.target.wants
dbus-org.freedesktop.nm-dispatcher.service | system
display-manager.service                  | system
docker.service                           | multi-user.target.wants
fstrim.timer                             | timers.target.wants
getty@tty1.service                       | getty.target.wants
gnome-keyring-daemon.socket              | sockets.target.wants
keyd.service                             | multi-user.target.wants
NetworkManager.service                   | multi-user.target.wants
NetworkManager-wait-online.service       | network-online.target.wants
nvidia-hibernate.service                 | systemd-hibernate.service.wants
nvidia-resume.service                    | systemd-hibernate.service.wants
nvidia-resume.service                    | systemd-suspend.service.wants
nvidia-resume.service                    | systemd-suspend-then-hibernate.service.wants
nvidia-suspend.service                   | systemd-suspend.service.wants
p11-kit-server.socket                    | sockets.target.wants
pipewire-pulse.socket                    | sockets.target.wants
pipewire-session-manager.service         | user
pipewire.socket                          | sockets.target.wants
remote-fs.target                         | multi-user.target.wants
sshd.service                             | multi-user.target.wants
systemd-userdbd.socket                   | sockets.target.wants
ufw.service                              | multi-user.target.wants
wireplumber.service                      | pipewire.service.wants
xdg-user-dirs-update.service             | default.target.wants

I'm using Docker too but I don't think it's affecting the routing:

> ip r get 8.8.8.8
8.8.8.8 via 192.168.50.1 dev enp5s0 src 192.168.50.2 uid 1000
    cache

Please post your complete system journal for the boot:
sudo journalctl -b | curl -F 'file=@-' 0x0.st
after some connection failures.

That would be here: https://0x0.st/KoeL.txt

This is after several failures (mtr, browser and curl).

ponsilio · 2025-08-27 11:05:45

Today I have been trying the kernel 6.16.1-arch1-1 and it seems to actually do the trick as gugah reported.
I will test some more.

Diskette · 2025-08-27 11:13:04

Happening to me too, on both laptop running Arch Linux and Steam Deck running CachyOS. I don't think it's kernel related though as CachyOS uses their own kernel and my laptop uses linux-lts.
I initially thought that it was my network since every single GNU/Linux distribution seemed to have domain resolving problems on my network specifically --- even using 1.1.1.1 and Google's 8.8.8.8. So I """"fix""""ed it by using Unbound and cache the DNS request beforehand, but in this scenario it does somewhat help... or that's just placebo.

Last edited by Diskette (2025-08-27 11:19:30)

xerxes_ · 2025-08-27 12:24:05

Patches from mainline kernels are backported to LTS kernels, if possible, so don't update both mainline and LTS kernel.

I use current LTS kernel - 6.12.43-1-lts and NetworkManager and didn't experience network problems (but don't use nvidia binary blob driver).
I tried gugah's script multiple times and everything was OK. Kernel 6.15.9 is also good.

I didn't tried 6.16 kernel line yet, because it have (or had?) ext4 serious bug (among other bugs).

If this network issues are kernel related maybe they are linked to some other factor: nvidia binary driver? - https://bbs.archlinux.org/viewtopic.php … 1#p2258331
Do you all experiencing network issues use nvidia binary driver with kernel < 6.16.1 or 6.12.43?

Last edited by xerxes_ (2025-08-27 12:32:00)

seth · 2025-08-27 13:44:58

I have not experienced this with 6.15.8 that I copied from the installation USB. I did test it couple of times back and forth so that the current and LTS always had failures while 6.15.8 did not. But it could be just a wonky coincidence.

For the LTS situation see https://bbs.archlinux.org/viewtopic.php … 9#p2258679 - the offending patch /has/ been backported into LTS.

route is fine, only "concerning" service would be ufw (but not plausible/likely) - and there're no other obvious problems in the journal.

idontgetit · 2025-08-27 14:17:22

xerxes_ wrote:

Do you all experiencing network issues use nvidia binary driver with kernel < 6.16.1 or 6.12.43?

Initially yes. However, a little later a laptop without any Nvidia HW or drivers started to show the same symptoms.

seth wrote:

the offending patch /has/ been backported into LTS.

Yeah, I get that now. The irregularity of this behavior got me thinking it was something else. Sorry for being obtuse, thank you for your patience.

only "concerning" service would be ufw (but not plausible/likely)

I also forgot to mention that these failures occurred even with ufw disabled and iptables rules flushed.

Arch Linux

#1 2025-08-25 19:32:57

[SOLVED] Seemingly random connection errors

#2 2025-08-25 20:52:15

Re: [SOLVED] Seemingly random connection errors

#3 2025-08-25 20:58:32

Re: [SOLVED] Seemingly random connection errors

#4 2025-08-25 21:08:27

Re: [SOLVED] Seemingly random connection errors

#5 2025-08-26 14:45:07

Re: [SOLVED] Seemingly random connection errors

#6 2025-08-26 14:51:57

Re: [SOLVED] Seemingly random connection errors

#7 2025-08-26 15:18:58

Re: [SOLVED] Seemingly random connection errors

#8 2025-08-26 15:27:42

Re: [SOLVED] Seemingly random connection errors

#9 2025-08-26 15:37:43

Re: [SOLVED] Seemingly random connection errors

#10 2025-08-26 16:12:09

Re: [SOLVED] Seemingly random connection errors

#11 2025-08-26 17:21:29

Re: [SOLVED] Seemingly random connection errors

#12 2025-08-26 18:26:03

Re: [SOLVED] Seemingly random connection errors

#13 2025-08-26 19:08:19

Re: [SOLVED] Seemingly random connection errors

#14 2025-08-26 19:52:00

Re: [SOLVED] Seemingly random connection errors

#15 2025-08-26 20:00:11

Re: [SOLVED] Seemingly random connection errors

#16 2025-08-26 20:05:06

Re: [SOLVED] Seemingly random connection errors

#17 2025-08-26 20:15:15

Re: [SOLVED] Seemingly random connection errors

#18 2025-08-26 20:19:17

Re: [SOLVED] Seemingly random connection errors

#19 2025-08-26 20:57:16

Re: [SOLVED] Seemingly random connection errors

#20 2025-08-27 10:30:35

Re: [SOLVED] Seemingly random connection errors

#21 2025-08-27 11:05:45

Re: [SOLVED] Seemingly random connection errors

#22 2025-08-27 11:13:04

Re: [SOLVED] Seemingly random connection errors

#23 2025-08-27 12:24:05

Re: [SOLVED] Seemingly random connection errors

#24 2025-08-27 13:44:58

Re: [SOLVED] Seemingly random connection errors

#25 2025-08-27 14:17:22

Re: [SOLVED] Seemingly random connection errors

Board footer