I don’t know why booting process isn’t completed. During boot, prompt says upower.service and dbus-broker.service failed.
1 Like
Did you edit/corrupt ? /usr/share/dbus-1/system.d/obex.conf
Nope . It was empty file. But now I have temporarily removed it by using
mv /usr/share/dbus-1/system.d/obex.conf /usr/share/dbus-1/system.d/obex.conf.bak
which has made the dbus-broker.service working but again the system doesn’t boot, blank screen appears with no messages of failures.
That file should not be empty.
$ dnf provides /usr/share/dbus-1/system.d/obex.conf
Updating and loading repositories:
Fedora 41 - x86_64 - Updates 100% | 150.5 KiB/s | 29.6 KiB | 00m00s
Repositories loaded.
bluez-obexd-5.81-2.fc41.x86_64 : Object Exchange daemon for sharing content
Repo : @System
Matched From :
Filename : /usr/share/dbus-1/system.d/obex.conf
bluez-obexd-5.81-2.fc41.x86_64 : Object Exchange daemon for sharing content
Repo : updates
Matched From :
Filename : /usr/share/dbus-1/system.d/obex.conf
You should consider whether the file is empty because the storage device ran out of space or is failing.
2 Likes
Can you post the output for
sudo rpm -Va | grep '^..[5]'
I’m quite certain that the storage device isn’t out of space and isn’t failing.
dnf provides /usr/share/dbus-1/system.d/obex.conf
Above command outputs cannot open log file read only filesystem. I don’t know enough linux such that I’m not understanding how it changed to read only filesystem as I previously changed obex.conf to obex.conf.bak 
Please post terminal output as searchable pre-formatted text using the </> button. This will help others with similar issue find this topic, so is a contribution to the community.
A filesystem may be mounted read-only if the system wasn’t shut down properly (which may result in a corrupt filesystem), but also a failing storage device or device out of space. It is likely that other files are corrupt.
If you are using df with btrfs the results are misleading – assuming you are using default btrfs, what does btrfs filesystem df / give?.
A Live USB Workstation Installer will allow you to use Gnome disks to check drive “health”.
1 Like
That’s a lot of changed files, did you have some sort of power outage during an update or similar?
2 Likes
Yeah. After I updated with dnf update, I forgot to restart and charge the battery. Battery depleted to zero, laptop didn’t shutdown properly with update unfinished and when I tried to boot, the error occurred. Is this contributed to the changed files? Restarting the system as soon as possible after finishing the update would’ve prevented or not? What should I do, now?
1 Like
If you have working network maybe try
sudo dnf reinstall mesa\*
reboot to see if it helps, if not maybe try the nuclear option and reinstall everything
sudo dnf reinstall $(rpm -qa --qf "%{NAME}\n") --skip-unavailable
1 Like
Thanks a lot @leigh123linux and @gnwiii for your incredible support. Didn’t need to go to the nuclear option:)
2 Likes
You should still verify the other changed files and reinstall the corresponding packages
rpm -qf --qf "%{NAME}\n" $(rpm -Va |grep '^..[5]' |awk '{print $3}')
The output gets stuck with no messages.
If you have a lot of changed files it can take 10 - 20 minutes to complete.
leigh@leigh-pc:~$ time rpm -qf --qf "%{NAME}\n" $(rpm -Va |grep '^..[5]' |awk '{print $3}')
texlive-texlive-scripts
minidlna
mock-core-configs
setup
livesys-scripts
fedora-repos
libdnf5
selinux-policy-targeted
unbound-libs
real 0m46.837s
user 0m35.696s
sys 0m10.904s
A lot of changed files are there and same package is outputted multiple times. When I tried to reinstall, conflicts occurs. There are different versions of same package.
Looks like I’ve installed pango from fedora source and anaconda. Doesn’t dnf treat anaconda packages different from the official fedora packages?
Ok, try (maybe backup anything important in /home , just in case).
sudo dnf install dnf-utils
sudo package-cleanup --cleandupes
Post
rpm -qa kernel\* upower\*