Hello Lemmy,
Im trying to troubleshoot a connection issue to my laptop which is connect using the buildin ethernet port. The connection drops sometimes for a few minutes and then im able to connect again.
This is the port being used:
k3s-alpine-lap-6:~# lspci -nn | grep -i ethernet
0000:01:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8111/8168/8211/8411 PCI Express Gigabit Ethernet Controller [10ec:8168] (rev 16)
Some information about the drivers being used:
k3s-alpine-lap-6:~# ethtool -i eth0
driver: r8169
version: 6.12.76-0-lts
firmware-version: rtl8168h-2_0.0.2 02/26/15
expansion-rom-version:
bus-info: 0000:01:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: yes
supports-priv-flags: no
Some information i found using dmesg:
[ 10.535868] Bridge firewalling registered
[ 29.478495] usb 3-1: USB disconnect, device number 2
[ 30.635978] atkbd serio0: Unknown key pressed (translated set 2, code 0xd8 on isa0060/serio0).
[ 30.635990] atkbd serio0: Use 'setkeycodes e058 <keycode>' to make it known.
[ 30.645244] atkbd serio0: Unknown key released (translated set 2, code 0xd8 on isa0060/serio0).
[ 30.645256] atkbd serio0: Use 'setkeycodes e058 <keycode>' to make it known.
[ 55.836493] EXT4-fs (sda): mounted filesystem f9742482-7e09-4460-a10d-81c5f5abaf23 r/w with ordered data mode. Quota mode: none.
[ 347.743005] Initializing XFRM netlink socket
[ 354.759146] eth0: renamed from tmp5fca0
[ 354.898482] eth0: renamed from tmp375a5
[ 354.930148] eth0: renamed from tmpaf7f3
[ 354.970131] eth0: renamed from tmpc035d
[ 356.883835] eth0: renamed from tmp68492
[ 518.679868] Key type trusted registered
[ 518.745372] Key type encrypted registered
[ 518.750554] device-mapper: uevent: version 1.0.3
[ 518.750614] device-mapper: ioctl: 4.48.0-ioctl (2023-03-01) initialised: dm-devel@lists.linux.dev
[ 599.682490] hrtimer: interrupt took 24965 ns
[ 832.545952] r8169 0000:01:00.0 eth0: Link is Down
[ 845.527825] r8169 0000:01:00.0 eth0: Link is Up - 100Mbps/Full - flow control off
[ 846.467804] r8169 0000:01:00.0 eth0: Link is Down
[ 848.089514] r8169 0000:01:00.0 eth0: Link is Up - 100Mbps/Full - flow control off
[ 2407.857541] r8169 0000:01:00.0 eth0: Link is Down
[ 2420.398450] r8169 0000:01:00.0 eth0: Link is Up - 100Mbps/Full - flow control off
Im not knowledgeable about this types of issues so im not even sure the driver is the issue.
Does someone have a simular experience or knows what could cause this issue i would love to hear it! Have a great day!
(Im sorry my post isn’t very specific about the issue but im not sure how i could better start troubleshooting this)
This isn’t uncommon. There’s a few things you can try:
https://www.baeldung.com/linux/realtek-ethernet-persistent-connection-issue
The other thing that is unusual is that you said when you connected a known good device it connected at 1Gbps but in your dmesg output it is only connecting at 100Mbps. Is it being plugged directly into your router/could you power cycle that device to see if maybe it is responsible for the link drop (also the cable could have some insulation damage and is picking up intermittent EMI).
Sounds like it could be a power saving setting kicking in.
Maybe relevent: https://unix.stackexchange.com/questions/729508/how-to-permanently-disable-eee-energy-efficient-ethernet-on-ethernet-card
It could be related to power saving settings and/or your power profile.
Some Q’s:
- Is your computer awake and in use when this happens?
- Are any of your USB devices disconnecting at the same time?
- I see your connection is 100Full and not 1Gb, is that because whatever it’s patched to is only 10/100?
I saw you tried to set power for the interface itself, but if this is a power issue, you probably want to disable power savings on your PCIe interfaces. Easiest way to do this is probably installing
powertop, and navigating over to the ‘tunables’ menu, and disabling power management for those interfaces, just to test and see if it fixes it.- Im not 100% certain but I assume yes. Im trying to set it up as a nas. I just did setup-alpine, rebooted it and placed it in my basement. It should have services consistently running on it like sshd, k3s, … which become unavailable/available every few minutes.
- Except for a keyboard that i used for troubleshooting no usb devices are connected. Should i attach an usb to see if it gets unmounted?
- I notices this aswell which i quite strange since it is connected using a cat 5.e cable which should be able to do 1G. The cable is connected to a gigabit switch so that also shouldn’t be the problem. Asfar as cabling and switching is concerned everything should be able to do 1G
If you need additional information let me know!
As additional troubleshooting step i connected a know working device to the cable and was able to get 1GB
Oh i just did
ethtool -I eth0and now it does show up as being able to do 1G.k3s-alpine-lap-6:~# ethtool -I eth0 Settings for eth0: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supported pause frame use: Symmetric Receive-only Supports auto-negotiation: Yes Supported FEC modes: Not reported Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised pause frame use: Symmetric Receive-only Advertised auto-negotiation: Yes Advertised FEC modes: Not reported Link partner advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Link partner advertised pause frame use: No Link partner advertised auto-negotiation: Yes Link partner advertised FEC modes: Not reported Speed: 1000Mb/s Duplex: Full Auto-negotiation: on master-slave cfg: preferred slave master-slave status: slave Port: Twisted Pair PHYAD: 0 Transceiver: external MDI-X: Unknown Supports Wake-on: pumbg Wake-on: d Link detected: yes Link Down Events: 2I’m not sure your rationale for picking Alipine for this use-case, but you might want to consider a more fleshed out distro for a NAS. Alpine is the BARE MINIMUM of an OS, meaning a lot of helpers that exist on other distros are not there for things like setting your power settings and link negotiations in certain cases. Sure the NIC driver is there, but all the nice tweaks for that specific module that exist in, say, Fedora Server aren’t going to be there, leading to this issue.
I’d honestly just throw FreeNAS, TrueNAS, or Unraid on this box and be done with it. They’ll have all the power settings and tweaks meant for a NAS in place, and then you won’t need to spend time hunting stuff like this down.
Very valid point.
Im actually running k3s on it so im not looking for any NAS specific OS. But if i picked a more general OS like fedora this issue might not have appeared. The reason i picked Alpine is since everything is running in containers i don’t need any fancy OS i just need to install k3s. But apperently Alpline can cause some issues :D
try running off a fedora usb stick long enough for this issue to have been triggered in alpine to confirm if this is the case without wiping out your installation.
Alpine is very specific in its use-case. Unless you know the exact package set you need to work with, it’s not suitable for general purpose use. It doesn’t even include glibc, for example 🤣 It’s meant to BARE as possible with a small footprint.
Thanks for your reply! This is a path i wanna look into. Just to confirm that im understanding correct running
/sbin/modprobe -i r8169 && sleep 1 && /usr/sbin/ethtool --set-eee eth0 eee offand then rebooting is what i should try right?I did the step i mentioned above but this wasn’t able to solve my issue :(
Additional information im on Alpine 3.22.3. If this is an Alpine specific issue im able to switch distros but i prefer not to if it isn’t needed.
