Failed update broke network... and possibly more

As the title says, I was on the rolling release of EndeavourOS (currently on LTS after this whole debacle) and my PC completely froze during the update process and had to force a reboot. Long story short, I had to do a lot of chroot to get to a working spot. Once I did, the network adapters were not seen and I had to manually enable the adapters on every reboot with ipconfig eno2 up then set a static IP to get internet. I fixed the issue by enabling systemd networking.

sudo systemctl enable systemd-network-generator.service
sudo systemctl enable systemd-networkd.service

Technically I’m pretty much working as expected at this point, but I cannot do anything with the firewall or adapters through the GUI. My networking tab in settings is completely blank and if I try to add an Ethernet connection, I get:

The name org.freedesktop.NetworkManager was not provided by any .service files

I can see network manager in the unit list but it’s masked and any attempt to unmask with

sudo systemctl unmask NetworkManager (+whatever service)

drops down the command line as if it’s successful but looking at the unit list again, it’s still in a masked state.

Uninstalling/re-installing and system updates have not fixed the issue.

Finally, I’m sure this is related, but the firewall will not launch, it immediately errors out and references IP tables possibly being borked. Again, updating this package and IP tables does not fix the issue.

Full output of firewalld.service status.

systemctl status firewalld.service
â—‹ firewalld.service - firewalld - dynamic firewall daemon
self.feed(b"", isFinal=True)
File “/usr/lib/python3.12/xml/sax/expatreader.py”, line 215, in feed
self._err_handler.fatalError(exc)
File “/usr/lib/python3.12/xml/sax/handler.py”, line 38, in fatalError
raise exception
xml.sax._exceptions.SAXParseException: :1:0: no element found

                                             During handling of the above exception, another exception occurred:

                                             Traceback (most recent call last):
                                               File "/usr/lib/python3.12/site-packages/firewall/core/fw.py", line 734, in start
                                                 self._start_failsafe()
                                               File "/usr/lib/python3.12/site-packages/firewall/core/fw.py", line 716, in _start_failsafe
                                                 self._start_load_stock_config()
                                               File "/usr/lib/python3.12/site-packages/firewall/core/fw.py", line 495, in _start_load_stock_config
                                                 self._loader_zones(config.FIREWALLD_ZONES)
                                               File "/usr/lib/python3.12/site-packages/firewall/core/fw.py", line 861, in _loader_zones
                                                 obj = zone_reader(filename, path, no_check_name=combine)
                                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                                               File "/usr/lib/python3.12/site-packages/firewall/core/io/zone.py", line 478, in zone_reader
                                                 raise FirewallError(
                                             firewall.errors.FirewallError: INVALID_ZONE: not a valid zone file: no element found: line 1, column 0

Jul 31 18:49:40 tron-endeavour firewalld[56053]: ERROR: INVALID_ZONE: not a valid zone file: no element found: line 1, column 0
Jul 31 18:49:40 tron-endeavour firewalld[56053]: ERROR: Failed to load full stock configuration. This likely indicates a system level issue, e.g. the firewall backend (nftables, iptables) is broken. All hope is lost. Exiting.
Jul 31 18:49:40 tron-endeavour firewalld[56053]: ERROR: Raising SystemExit in run_server

I’m thinking I may have more issues beyond this, but everything seems to be working besides what’s listed above. The reason I may have more issues is the output of systemctl --all --no-pager

Which shows the following as “not-found”
boot.automount
boot.mount
home.mount
run-credentials-systemd\x2dresolved.service.mount
run-credentials-systemd\x2dsysctl.service.mount
sysroot.mount
auto-cpufreq.service
NetworkManager.service
nslcd.service
plymouth-quit-wait.service
plymouth-quit.service
plymouth-start.service
syslog.service
system76-power.service
systemd-quotacheck.service
tlp.service
tuned.service
display-manager.target

As stated, I seem to be running fine and can even update the OS without further issue on LTS. Home drive works, network is fine (outside of NetworkManager and Firewalld) and Nvidia drivers installed. I technically have a dual boot system (if this info matters) with Windows but I pulled the Windows OS and installed EndeavourOS on a totally separate drive and opted to change the boot in BIOS to switch between them. Goal being on Linux full time and decom the Windows HD.

I’m kinda new to Linux (but worked on specialized servers) and have basically exhausted my google searches. Any help would be much appreciated.

Welcome @genghis_tron :wave: :sunglasses: :enos_flag: (I chuckled when I read that username :smile:)

I respect and admire your perseverance if you want to persist with trying to fix these issues.

As it’s not entirely clear what was done when you “had to do a lot of chroot to get to a working spot”, and the scope of the issue is unclear, personally, I would just backup important data and start fresh. At least then I can rest assured, everything is as it should be, particularly as there are some security implications with your current issues.

That’s just my 2cents. Someone else with more Linux wisdom than I (and there are many here), may have something more relevant to your current issues to offer. Best of luck!

(if you do decide to re-install fresh, let us know beforehand so someone doesn’t unnecessarily spend a heap of time offering no longer relevant suggestions)

1 Like

Actually, re-installing sounds a good plan, unless it’s a complicated system or something else.

If you want to repair it, you start from pacman log, to find where the crash left the upgrade process.

1 Like

I figured my chroot comment would get some pushback :grinning:

Regrettably I wish I had documented everything I did. This was my first time using chroot so initial setup took a bit to figure out. I mounted my /mnt and /mnt/efi partitions (I didn’t set up any complex partitions during initial install) and attempted pacman -Syu on the rolling distro and really didn’t get anywhere, I do recall NVIDIA drivers being an issue at this point, so I removed official and went to nouveau which didn’t help.

I decided to move to LTS and actually got some boot information on screen and had DBUS errors so I installed DBUS broker, which helped some, but I remember stumbling on a thread to remove broker and install daemon, so that’s what I did and managed to boot further.

The last bit was getting stuck on “reached target graphical interface” on boot which startx command would fail on. I removed nouveau and installed the official drivers and got a little further. Lastly I installed SDDM to get a display manager which allowed me to login. There’s probably some… or a lot that I’ve simply forgotten that I attempted, but I do remember those problems specifically.

After that is where the network problems listed in my original post came in. I can certainly re-install if that’s the consensus.

Appreciate your feedback. I don’t have anything real complex or important. This is my gaming rig and I just want to be off Windows at this point. The most “complex” thing I have is an LVM group/cache which a 8TB rust spinner and NVME are in to boost read/write. It was more of a test project than anything. It seems to be working as expected but I have another 4TB NVME drive coming so I might just abandon it altogether.

Before the crash I had finally set up my OS just the way I wanted it, so that was the main reason for being re-install adverse. That, and it was an opportunity to learn more about fixing broke stuff that I could encounter in the future.

1 Like

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.