You are not logged in.
It was working a few weeks ago but I noticed slow transfers between my computers yesterday.
The router supports 1gbit (and it was working in the past).
I have tried the following:
- Rebooting the router
- Rebooting the PC
- A different ethernet cable
- A different port on the router (where another PC works fine at 1gbit)
Only one PC is affected despite both running arch.
This is the device:
[ 5.986939] alx 0000:05:00.0 eth0: Qualcomm Atheros AR816x/AR817x Ethernet [e0:d5:5e:23:6c:9f]
[ 6.032581] alx 0000:05:00.0 enp5s0: renamed from eth0
I'd imagine it can only be a kernel update that's caused a problem. Here's the output from ethtool:
ethtool enp5s0
Settings for enp5s0:
Supported ports: [ TP ]
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: Not reported
Advertised pause frame use: No
Advertised auto-negotiation: No
Advertised FEC modes: Not reported
Speed: 100Mb/s
Duplex: Half
Port: Twisted Pair
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
MDI-X: Unknown
It's only running at half duplex 100Mbit but it is showing as supporting 1gbit (and this speed was working in the past).
Any ideas what I can do to fix it? I tried configuring NetworkManager to connect 1000mbit full duplex, doing so did give me full duplex but still only 100Mb/s.
Last edited by Tom B (2019-01-12 15:00:48)
Offline
Ok, a pacman -Syu and a reboot seems to have fixed it. I'm not sure what package it was as there was no kernel update
Offline
What packages were updated? It may help figure out what caused the problem in the first place, and help someone with the same problem.
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Offline
Looking at the packages, I'm not convinced that was the fix.
[2019-01-12 14:38] [ALPM] transaction started
[2019-01-12 14:38] [ALPM] upgraded libutil-linux (2.33-3 -> 2.33.1-1)
[2019-01-12 14:38] [ALPM] upgraded util-linux (2.33-3 -> 2.33.1-1)
[2019-01-12 14:38] [ALPM] upgraded ark (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded baloo-widgets (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded docker (1:18.09.0-2 -> 1:18.09.1-1)
[2019-01-12 14:38] [ALPM] upgraded kio-extras (18.12.0-2 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded phonon-qt5 (4.10.1-3 -> 4.10.2-1)
[2019-01-12 14:38] [ALPM] upgraded dolphin (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded doxygen (1.8.14-1 -> 1.8.15-1)
[2019-01-12 14:38] [ALPM] upgraded firefox (64.0-1 -> 64.0.2-1)
[2019-01-12 14:38] [ALPM] upgraded signon-kwallet-extension (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded kaccounts-integration (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded kate (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded kcontacts (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded konsole (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded lib32-libxslt (1.1.32+3+g32c88216-1 -> 1.1.33-1)
[2019-01-12 14:38] [ALPM] upgraded libkexiv2 (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded libkipi (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded libplacebo (0.6.0-2 -> 1.7.0-1)
[2019-01-12 14:38] [ALPM] upgraded qca (2.1.3-2 -> 2.1.3-3)
[2019-01-12 14:38] [ALPM] upgraded okular (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] warning: /etc/pacman.d/mirrorlist installed as /etc/pacman.d/mirrorlist.pacnew
[2019-01-12 14:38] [ALPM] upgraded pacman-mirrorlist (20181205-1 -> 20190109-1)
[2019-01-12 14:38] [ALPM] warning: /etc/makepkg.conf installed as /etc/makepkg.conf.pacnew
[2019-01-12 14:38] [ALPM] upgraded pacman (5.1.1-1 -> 5.1.2-2)
[2019-01-12 14:38] [ALPM] upgraded python-docker (3.6.0-1 -> 3.7.0-1)
[2019-01-12 14:38] [ALPM] upgraded rubygems (2.7.7-1 -> 3.0.2-3)
[2019-01-12 14:38] [ALPM] upgraded ruby (2.5.3-1 -> 2.6.0-4)
[2019-01-12 14:38] [ALPM] upgraded spectacle (18.12.0-1 -> 18.12.1-1)
[2019-01-12 14:38] [ALPM] upgraded steam (1.0.0.56-1 -> 1.0.0.59-1)
[2019-01-12 14:38] [ALPM] upgraded vagrant (2.2.2-1 -> 2.2.2-2)
[2019-01-12 14:38] [ALPM] upgraded virtualbox (6.0.0-2 -> 6.0.0-3)
[2019-01-12 14:38] [ALPM] upgraded vlc (3.0.4-7 -> 3.0.6-1)
[2019-01-12 14:38] [ALPM] upgraded x11vnc (1:0.9.15-1 -> 1:0.9.16-1)
As far as I can tell, there's nothing here that should affect network connections. I did boot into Windows first to check the speed there (it was 1gbit as expected) before rebooting into arch, but I don't see how that would have affected anything unless it had correctly initialized the hardware or something, but it seems unlikely. If it happens again I'll see if I can work it out.
Last edited by Tom B (2019-01-12 20:11:37)
Offline
Thanks for posting some more info. You're right, nothing in that list jumps out at me as the cause either.
A full journal from an affected session may provide more insights (e.g. `journalctl -b -1`), but this wouldn't be the first instance I've seen of something behaving differently depending on which OS was previously booted. The next time it happens you may want to use ethtool to workaround the problem, setting the duplex to full, and speed to 1000, and see if that works.
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Offline
Auto-negotiation: off
This is the most suspicious setting, this should be enabled. I have no idea why it wasn't, though.
| alias CUTF='LANG=en_XX.UTF-8@POSIX ' |
Offline