i486/i686/pentium4 on 15 systems.
:-)
]]>systemd got stuck in a unresolved dependency on libip4tc.so.0/libip4tc.so.2, so I had to force rebuild iptables.
I pushed i686 and pentium4 packages accordingly.
The i486 got stuck in 5.1.x, dunno why. Most likely the few build slaves are still building unnecesary things, so the linux
kernel and systemd fell behind.
This is a bad bug. Now I know why my remote Arch 32 box did not come back up after the update I did a couple of days ago. Can somebody please post here when the bug is fixed as I won't bother driving over to it until then.
Not a fix, but I do have two workarounds:
1) create a cron entry for root (assuming you're running cron) that has "@reboot ip link set enpxxxxx up" where enpxxxx is the name of your interface
or
2) Even though you probably have a static IP address, enable dhcpcd:
# systemctl enable dhcpcd
# systemctl start dhcpcd
Either of these will bring the interface up at boot time.
]]>● systemd-networkd.service - Network Service
Loaded: loaded (/usr/lib/systemd/system/systemd-networkd.service; disabled; v>
Active: inactive (dead)
Dunno why it says disabled. I only updated today fwiw, so I'm actually still running the old 5.1.x kernel at present. I also see nothing in journalctl under any sensible unit names I can guess at, and netctl status looks good although it does say my logs have been rotated today, so that might explain some of this oddness.
]]>Can reproduce this. Quite annoying. I don't know if this is a kernel, a systemd-networkd or whatever bug..
systemctl status systemd-networkd says:
ens3: Could not bring up interface: Invalid argument
I see the "Invalid argument" in 5.1.16 too, but the interface still comes up.
]]>(abaumann wonders what kernels are supported and tested with systemd)
]]>