/etc/fstab drops it into emergency mode, a damaged filesystem needs fsck, or a new kernel fails to start. Open the VNC console, read the last messages, then fix /etc/fstab, repair the filesystem from rescue mode, or pick the previous kernel in the GRUB menu.
Before you start
- On a VPS, open the VNC console to see the screen. For repairs on a system that can’t start at all, use rescue mode. On a dedicated server, open a ticket for KVM/IPMI access.
- Emergency mode asks for the root password. If you never set one, reset the root password or work from rescue mode.
- If you have backups, note which one is the latest before you change anything.
Step 1: Read what the console says
In emergency mode, log in as root and read the reason:
Cause 1: A wrong line in /etc/fstab
/etc/fstab lists the disks to mount at boot. A typo, a wrong UUID or a removed disk makes systemd wait, give up and stop in emergency mode.
Check (in emergency mode):
1
Make the root filesystem writable
2
Correct or disable the bad line
# in front of the line findmnt reported, or fix its UUID (list real ones with blkid). For extra disks that may be missing, add nofail so the boot continues without them:3
Check and reboot
Cause 2: The filesystem needs a check
After a crash or a forced power-off, a filesystem can have errors that the automatic check at boot won’t fix on its own. Fix at the initramfs prompt (Ubuntu and Debian often stop here with the device name in the message):-y answers “yes” to every repair.
Fix from rescue mode if there’s no prompt, or for XFS (the default filesystem on AlmaLinux and Rocky):
1
Start rescue mode and find the disk
Start rescue mode, connect, then list disks and filesystems:
2
Repair the filesystem without mounting it
- ext4 (Ubuntu / Debian)
- XFS (AlmaLinux / Rocky)
3
End rescue mode
End the rescue session in the panel. The server boots normally.
lsblk -f shows LVM2_member), run vgchange -ay first and repair the logical volume, for example /dev/mapper/almalinux-root.
Cause 3: A new kernel doesn’t start
A kernel panic straight after an update often means its initramfs (the small startup image) is missing or incomplete. Fix: boot the previous kernel, then rebuild the initramfs.1
Open the GRUB menu
Reboot from the panel and watch the VNC console. On Ubuntu the menu is hidden: hold Shift (BIOS) or press Esc (UEFI) as the server starts. Choose Advanced options for Ubuntu and an older kernel. On AlmaLinux and Rocky, press an arrow key when the kernel list appears to stop the countdown, then pick the older entry.
2
Rebuild the initramfs
- Ubuntu / Debian
- AlmaLinux / Rocky
3
Reboot into the newest kernel
Reboot without choosing anything. If it fails again, keep using the older kernel and look for a free-space problem in
/boot (df -h /boot).Cause 4: The bootloader is broken
Agrub rescue> prompt means GRUB can’t find its files, for example after a partition change. Start rescue mode, mount the disk and copy your important data off first. Then open a ticket or rebuild the server. Check that the boot type matches how the system was installed.
How to check it worked
The console shows a normal login prompt, and after login:Frequently asked questions
Emergency mode says “the root account is locked”. What now?
Ubuntu doesn’t set a root password by default, so emergency mode can’t open a shell. Use rescue mode, mount the disk and edit/etc/fstab there (/mnt/etc/fstab once mounted on /mnt).
Will fsck delete my files?
It repairs the filesystem’s structure. Damaged files may end up inlost+found or be lost, which is why a backup comes first when you can.
The server boots but SSH doesn’t work. Is that the same problem?
No. See SSH connection refused and SSH connection timed out.Related guides
VPS and dedicated server guides
Every guide in this section.
Read-only file system errors
When the disk stops accepting writes.
Mount a new disk
Safe fstab entries with nofail.
Read logs with journalctl
Find the boot error in the log.
Rescue mode
Repair a VPS that won’t start.
Checked against the official systemd source for the emergency mode message, the fstab, findmnt, e2fsck and xfs_repair manuals on 4 November 2026;
findmnt --verify with a bad fstab line, e2fsck and xfs_repair on unmounted test filesystems and the update-initramfs and dracut options were checked on Ubuntu 24.04 LTS. Booting, the GRUB menu and LVM steps were not tested.Need Extra Help?
If you encounter any issues, our support team is ready to assist:- Live Chat: Quick assistance via our website.
- Support Ticket: Open a Ticket
- Discord: Join our Community
- Email: support@freakhosting.com
Last Updated: November 2026