I think we should focus on a KISS solution: keep it short and simple
@jakfrost : Am I right that your current goal is to create a working terminal on the damaged system to then do further investigation on the terminal? Therefore, a simpler implementation of the chroot approach above? I do not know much about the rescue option because I have not yet needed it, but if the rescue provides only the same outcome as the chroot approach (which is a terminal that uses the damaged system as root), I think the easiest and fastest way is to use any live system to backup the data using the GUI (the btrfs drive seems to be ok), and then simply install a new Fedora. Even if the logs contain sufficient indication, it is likely that complexity even increases in subsequent steps.
@perryt :
→ Do you have backups of the data that is stored on the damaged Fedora? You do not need backups of the operating system itself. I mean only backups of your own data (private data, business data, documents, pictures, videos, and so on).
→ Would it be a problem to delete your Fedora and install a new Fedora? If this approach is ok for you:
You can do this with a live system, which you can create with the media writer (for this, use the mediawriter, and create a Fedora Workstation 36 (not Server!):
If you have not backups of all your data, start the live system, click on Activities
on the top left and open files
. Then the file manager opens. On the left, click on Other locations
. Then, check out the drives on this computer
: One of them is your old Fedora, which contains the following folders: home
, root
, timeshift
.
Search in these three folders for the data you want to backup. Use an additional USB storage and backup this data (it has to be an USB storage other than the one with the live system; maybe you can borrow a USB storage for a day if you have no second one?). You can do the backup activities in the file manager. Check out on another system if the data is backed up on the USB storage! Now you have a backup:
If you have backups of all your data, start the live system, click on Activities
on the top left and open install to hard drive
. In installation destination
, use automatic
, done
and then reclaim space
. Then, delete all
and then reclaim space
. This will delete all remaining data on this system! Of course, you can also do this manually instead of automatic.
If all your private files are stored on the home partition, you can re-use this partition without formatting it (installation destination → custom → mount+format boot, mount+format root, only mount home without format). In this case, you can create a new Fedora installation without deleting your private files. However, in this situation, I suggest to not try anything without a backup!
Hope this helps.
Backup: MegaNZ backup - i will backup, and “full disk space”.
I am megaNZ and i is try backup.
So, i am usb in fedora workstation 36 (not server)!!
…
Do They can server rescue option ?
…
Sounds good. Do the backup, and when you have verified that the backup contains all your data, simply delete and re-install Fedora, as described above.
Sorry. I do not understand this question. But re-installing Fedora will not need a rescue option. You will need the workstation, not server.
I think this approach will practically be easier and faster for you.
Sorry. I am aphasia!! Sorry.
Do They can server rescue option ?
Do fedora wizards can server rescue option? I am server rescue option that instructions. I am usb and server rescue option .
No worries. We get this done somehow. Can you maybe try to re-phrase the question?

Do fedora wizards can server rescue option? I am server rescue option that instructions. I am usb and server rescue option .
Concerning the wizard: no, the installation wizard does not have a rescue option. We can only use it to delete Fedora and then, install Fedora again. But you can re-use some existing partitions (see above): this could includes home
and timeshift
but not root
. But make a backup before that anyway!
The wizard in Fedora’s installation is one of the reasons why I think this approach is the easiest and fastest for you after you have made backups. Also, you seem already familiar with these types of processes (installation, using/mounting partitions, and such).
I think that there is no wizard for rescue other than what you have in the Server’s rescue option. But I have never used one. @jakfrost can you help with that?
Hello,
Sorry for the absence, life get’s in the way, no?
First, I have not used the rescue option since F27 I think. At that time I don’t remember explicitly having to go through any particular steps at the command line, just pick it and then it did it’s magic. If it doesn’t work like that or similar, I would abandon spending too much time on it.
Re-installation is much easier. For backup’s BTRFS snapshots are an option I have used before, and did that from the command line using the Live USB to boot the system then just snapshot the partitions I wanted to backup, then re-install and snapshot send/receive to put data back, or just copy it. If you use snapshots, remember to make them RW not just read (which is default) since to use the send/receive option you need the backup snapshot to be RW. The installation wizard won’t rescue your system, but is very reliable at installing a new system over top. I have also used rsync in the past as a nifty backup tool, but the BTRFS filesystem already has snapshot’s built in and they’re faster by far. There is nothing needed to install to do a snapshot, simply use the command sudo btrfs snapshot /home <nameofbackupsubvol>
to make a backup of your /home (since it is already a btrfs subvolume, if you used default disk layout for original installation).
In my case my /etc/fstab looks like this for my home mount and the backup subvolume …
UUID=ae8acfae-a997-440d-a2ac-831ec4015694 /home btrfs subvol=home 0 0
UUID=ae8acfae-a997-440d-a2ac-831ec4015694 /home/snapshots btrfs subvol=home-bak 0 0
I know the backup is to the same drive but this is an easy way to back up if your drive is capable. Plus, a snapshot can also be moved using the mv
command.
do you is /etc/fstab is liveuser usb? sudo mount /dev/sda3 /mnt/home?
@jakfrost I fear a high likelihood that the btrfs snapshot gets destroyed during the installation. Especially if the automatic partitioning is used, I am not sure if the installation will set up a completely new btrfs (deleting the old one, including its snapshots). In terms of custom partitioning, Perry seems to have a very good background knowledge but I see a risk in the interpretation of the texts he faces during the installation destination
steps. This is why I suggested to make an additional backup despite the possibility to re-use the home partitions. Therefore, I think his idea of using his cloud storage for the transition makes most sense? Although using both makes sense as well.
@perryt just to ensure that you do not loose data, and to create as much security as possible, I suggest to implement your megaNZ cloud solution in either case (because you suggested it, I assume there are no privacy issues with it). Of course Stephen’s idea can create additional security, and if the btrfs file system remains after the installation, Stephen’s solution will be much faster and easier in getting access to the data on the new Fedora. Maybe re-using the home partition works as well. But we should be prepared for everything

do you is /etc/fstab is liveuser usb? sudo mount /dev/sda3 /mnt/home?
i am help!
i am sudo vi /etc/fstab – liveuser usb
[liveuser@localhost-live etc]$ sudo systemctl daemon-reload
[liveuser@localhost-live etc]$ sudo mount -a
mount: /mnt/home: /dev/sda3 already mounted on /mnt/home.
dmesg(1) may have more information after failed mount system call.
mount: /mnt/home/snapshot: mount(2) system call failed: No such file or directory.
dmesg(1) may have more information after failed mount system call.
less fstab
vartmp /var/tmp tmpfs defaults 0 0
UUID=6fec64c0-a806-464e-9e7e-a72a403ba06e /mnt/home btrfs subvol=home 0 0
UUID=6fec64c0-a806-464e-9e7e-a72a403ba06e /mnt/home/snapshot btrfs subvol=home-bak 0 0
do i am chroot /mnt/root/root?
i am dnf search snapshot
========================================= Name & Summary Matched: snapshot =========================================
baresip-snapshot.x86_64 : Snapshot video filter using libpng for baresip
rsnapshot.noarch : Local and remote filesystem snapshot utility
texlive-snapshot-doc.noarch : Documentation for snapshot
============================================== Name Matched: snapshot ==============================================
golang-github-containerd-fuse-overlayfs-snapshotter.x86_64 : Fuse-overlayfs plugin for rootless containerd
golang-github-containerd-fuse-overlayfs-snapshotter-devel.noarch : Fuse-overlayfs plugin for rootless containerd
golang-github-containerd-stargz-snapshotter.x86_64 : Fast container image distribution plugin with lazy pulling
golang-github-containerd-stargz-snapshotter-devel.noarch : Fast container image distribution plugin with lazy
: pulling
golang-github-containerd-stargz-snapshotter-estargz-devel.noarch : Fast container image distribution plugin with
: lazy pulling
perl-Test-Snapshot.noarch : Test against data stored in automatically-named file
rust-cranelift-codegen+regalloc-snapshot-devel.noarch : Low-level code generator library
stargz-snapshotter.x86_64 : Fast container image distribution plugin with lazy pulling
texlive-snapshot.noarch : List the external dependencies of a LaTeX document
============================================ Summary Matched: snapshot =============================================
which snapshot?
[liveuser@localhost-live etc]$ which snapshot
/usr/bin/which: no snapshot in (/home/liveuser/.local/bin:/home/liveuser/bin:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin)
@perryt My suggestion is to only do the backup with your megaNZ.
Not btrfs snapshots.
If you have done your backup, verify that all data is backed up.
Then, re-install Fedora as described above.
This way we focus on GUIs, wizards and steps you are already used to.
MegaNZ backup – i afraid MegaNZ Backup – i am MegaNZ “full disk space”.
MegaNZ – i locate download to the not link.
i am not app
[liveuser@localhost-live etc]$ dnf search meganz
Last metadata expiration check: 1 day, 0:35:51 ago on Thu 02 Jun 2022 09:47:12 AM EDT.
No matches found.
i am software “not application found.”
There is no megaNZ available with dnf. I assumed you have a cloud account there because you mentioned it for doing backups. Therefore, I assumed you already use this service or that you have experience with it. I do not know if you need an app for megaNZ.
If you have files that need to be backed up, you have to find a solution for that. You can use external storages (although we need one dedicated USB stick for re-installing Fedora) or a cloud storage provider.
If you are running a live system, you cannot install much because installed software has to be stored in the RAM.
I suggest to avoid the command line for now. Given the issues we already had, this can be very dangerous.
do fedora they is new OS? I will intend hardware OS. The Fedora is asking people is new OS?
start new OS , and MegaNZ backup … do i doing restart backup …
thank you py0xc3, jakfrost, etc …
i am 1980’s programmers Unix – VAX, Apollo, Sun, etc – instead Linux Fedora.

do fedora they is new OS? I will intend hardware OS. The Fedora is asking people is new OS?
Can you re-phrase that? I assume you mean with hardware that you want to install the system permanently on the disk instead of just using the live system? This will work after you re-installed Fedora.

start new OS , and MegaNZ backup … do i doing restart backup …
You start the live system you already have. Then you use the live system to do the backup. You can upload your data to a cloud, or you can copy/backup it on an external drive. Then you use on the live system the Install button to install a new Fedora.
Hello @perryt ,
Yes you can install Fedora Linux onto bare metal hardware.
As @py0xc3 mentioned, it is likely easier to backup data (if you have it to back up) just by copying it to another storage device if available. Then follow the default installation using the Fedora media writer to make the installation media.
As for the BTRFS snapshot of your /home subvolume, it would still exist after installation, but you would have to create a mount point for it after the install finished to use it to copy from, or conversely leave it unmounted and use the send/receive commands on snapshots to do restoration of /home. Although it is the fastest to physically do, it is a bit more complex if you’re not familiar with BTRFS, the learning curve is a bit steep if you’re doing it and trying to actually use your device.
I ran Solaris 10 for a bit of time on my home machine, still have the disks, also used Sun’s Java Desktop, which was using a linux kernel.
There are similarities between the 'nixes, just same but different in many cases.
So, I am just going to ask, since this thread has meandered a bit. This problem started after an update of a working and installed (on bare metal) Fedora Linux system, correct? If so, @py0xc3 is there a rollback capability in dnf to revert an upgrade? I am spoiled, I use Silverblue and can rollback to my previous deployment at the boot menu. Haven’t used WS since F27.

If so, @py0xc3 is there a rollback capability in dnf to revert an upgrade?
There is no supported way of reverting to 35. Unsupported ways are more complicated (and error-prone) than doing a backup and re-install Fedora. Given the development of the thread, I think this is the easiest way to solve the original problem. My goal is to focus on types of operations/processes Perry is already used to, or that are wizard-based. I hope that way helps him to get his machine back to work.

There is no supported way of reverting to 35. Unsupported ways are more complicated (and error-prone) than doing a backup and re-install Fedora.
Darn, that’s what I thought. The thread here points to two very prominent pain points with Linux Distro’s in the general sense, recovery from system changes and decent rescue method. By running Fedora Linux Silverblue I can avoid/mitigate the first (recovery from an upgrade) by rolling back to my previous deployment with a single command rpm-ostree rollback
. However, there is no good way to recover an existing system cleanly by a user, without having to know sysadmin level commands.
[Edit] @perryt , The more I think about it the more I am convinced Silverblue would be a good idea for you to help avoid future potential similar issues. I never am left without a working system since I have begun using it, even with rawhide for a good portion of the time since F27. It is not much different than Workstation mostly a bit of a POV shift on usage. Most of the tools and all of the same skill sets to use effectively.
i am not applications (not BitWarden & not Mozilla) & KDE & OS (not websites 0.0.0.0 & jobs crontabs) !! I am files backup MegaNV.
i am live system USB Fedora 36. I am live system USB gets reinstalling Fedora 36.