Symptom: your Ubuntu machine boots to a black screen that says "Entering emergency mode. Exit the shell to continue.", followed by red errors about the system bus and /bin/bash. This guide explains what those messages actually mean, how to find the real cause, and how to repair the most common one: a failed root filesystem check on an LVM volume.
Tested scenario: Ubuntu 26.04 LTS, ext4 root on LVM (/dev/mapper/ubuntu--vg-ubuntu--lv). The same steps apply to Ubuntu 25.10 and other dracut-based distributions such as Fedora.
Quick fix (TL;DR)
If your boot log shows [FAILED] Failed to start systemd-fsck-root.service, the root filesystem check failed and Ubuntu refused to mount it. At the # prompt:
journalctl -u systemd-fsck-root.service --no-pager | tail -40
blkid /dev/mapper/ubuntu--vg-ubuntu--lv
fsck.ext4 -f /dev/mapper/ubuntu--vg-ubuntu--lv
rebootBack up first if the data matters. If the disk is physically failing, running a repair can make things worse. Details below.
The error messages
On screen you will typically see some combination of these lines:
Failed to connect to system scope bus via local transport: No such file or directory
Generating "/run/initramfs/rdsosreport.txt"
Entering emergency mode. Exit the shell to continue.
Type "journalctl" to view system logs.
Press Enter for system maintenance
(or press Control-D to continue):
sulogin: failed to execute /bin/bash: No such file or directory
#Scrolling further up (or rebooting and watching closely) reveals the line that matters:
[FAILED] Failed to start systemd-fsck-root.service - File System Check on /dev/mapper/ubuntu--vg-ubuntu--lv
[DEPEND] Dependency failed for sysroot.mount - /sysroot.
[DEPEND] Dependency failed for initrd-root-fs.target - Initrd Root File System.
[DEPEND] Dependency failed for initrd-parse-etc.service - Mountpoints Configured in the Real Root.Why the red errors are not the real problem
You are not in your installed Ubuntu system yet. You are inside the initramfs, a tiny temporary system that loads before your real root filesystem. Since Ubuntu 25.10, Ubuntu desktop builds this initramfs with dracut instead of initramfs-tools (OMG! Ubuntu). The rdsosreport.txt message is dracut's signature.
That minimal environment does not include D-Bus or Bash, so:
- "Failed to connect to system scope bus" just means there is no D-Bus in the initramfs. It is a side effect, not the cause.
- "sulogin: failed to execute /bin/bash" means Bash is not in the initramfs. The emergency shell falls back to a basic
sh, which is the#prompt you see. It works fine for repairs.
Pressing Ctrl+D will only retry the same boot and land you back here. The fix is to deal with whatever stopped the root filesystem from mounting.
Finding the real cause
Emergency mode means the root filesystem could not be mounted. The usual causes are:
| What the log shows | Likely cause |
|---|---|
Failed to start systemd-fsck-root.service | Filesystem errors (often after a power cut or forced shutdown), or a failing disk |
Timed out waiting for device | Disk not detected, missing storage driver in initramfs, or wrong UUID in GRUB/fstab |
| No password prompt on an encrypted disk | Missing LUKS/cryptsetup support in the initramfs |
This guide covers the first row, which is the one shown in the screenshots above.
Step-by-step fix: repair the root filesystem
Step 1: Read the actual fsck error
The emergency shell may not include less, so pipe through tail instead:
journalctl -u systemd-fsck-root.service --no-pager | tail -40Messages such as "UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY" confirm filesystem corruption. I/O errors or "Buffer I/O error on device" suggest the disk itself is failing. If you see I/O errors, skip ahead to the backup warning in Step 3.
Step 2: Confirm the LVM volume is active and check its filesystem type
lvm lvscan
blkid /dev/mapper/ubuntu--vg-ubuntu--lvlvscan should list the volume as ACTIVE. If it shows inactive, activate it with lvm vgchange -ay. The blkid output includes TYPE="ext4" (Ubuntu's default), xfs, or btrfs.
The double hyphens in ubuntu--vg-ubuntu--lv are not a typo. Device-mapper escapes a hyphen inside a name by doubling it. The same volume is also reachable as /dev/ubuntu-vg/ubuntu-lv.
Step 3: Back up before repairing, if you can
fsck changes the filesystem in place. On a healthy disk with minor corruption it is safe and routine. On a physically failing disk, the extra reads and writes can turn a recoverable situation into lost data. If anything on the machine is not backed up, boot from an Ubuntu live USB first and copy your important files off. For badly damaged disks, clone the drive with ddrescue first.
Step 4: Run the repair
For ext4, the volume is not mounted at this point, so it is safe to check:
fsck.ext4 -f /dev/mapper/ubuntu--vg-ubuntu--lvAnswer y to the proposed fixes. If there are hundreds of prompts, re-run with -y to accept them all:
fsck.ext4 -fy /dev/mapper/ubuntu--vg-ubuntu--lvRun it again until it reports the filesystem as clean with no further changes. fsck's exit code tells you the outcome:
0: no errors.1: errors were corrected.4or higher: errors remain uncorrected. Do not reboot and hope; investigate further.
You can check the code with echo $? right after fsck finishes.
Step 5: Reboot
rebootIf the filesystem is now clean, Ubuntu should boot normally.
If fsck.ext4 is not found
If the shell replies fsck.ext4: not found, the initramfs was built without the filesystem checker. That is why the automatic check failed, and your disk may be perfectly healthy. Repair from a live USB instead:
- Boot an Ubuntu 26.04 live USB and choose Try Ubuntu.
- Open a terminal and activate LVM, then run the check:
sudo vgchange -ay sudo fsck.ext4 -f /dev/mapper/ubuntu--vg-ubuntu--lv - Reboot into your installed system and regenerate the initramfs so the checker is included next time:
sudo dracut --force --regenerate-all
After you boot: checks you should not skip
Getting back to the desktop does not mean the problem is solved. Find out why the filesystem broke.
Check disk health
sudo apt install smartmontools
lsblk -d -o NAME,MODEL,SIZE
sudo smartctl -a /dev/nvme0n1 # or /dev/sda for SATA disksWarning signs include reallocated sectors, pending sectors, media errors, or a health assessment other than PASSED. If you see any of these, replace the drive soon and keep backups current.
Review errors from the failed boot
journalctl --list-boots
journalctl -b -1 -p err --no-pagerLogs from inside the initramfs are not always kept in the persistent journal, so the previous boot may show little.
Prevent it happening again
FAQ
Is "sulogin: failed to execute /bin/bash" a sign that Bash was deleted?
No. It refers to the initramfs, which does not ship Bash. Your installed system's /bin/bash is untouched.
Will fsck delete my files?
fsck fixes filesystem structures. Fragments of files it cannot reattach are placed in /lost+found. On mild corruption you typically lose nothing, but there is no guarantee on a damaged disk, hence the backup advice.
My root filesystem is XFS or Btrfs. What do I run?
For XFS, use xfs_repair on the unmounted device. For Btrfs, start with the read-only btrfs check. Do not use btrfs check --repair without a backup; the Btrfs developers themselves warn it can cause further damage.
Can I just boot an older kernel from GRUB?
That helps when a kernel or initramfs update is the cause. It does not help with filesystem corruption, because every kernel mounts the same broken filesystem.
