Skip to main content
A rescue ISO starts a separate Linux environment without reinstalling the VPS. Use it to inspect a system that will not boot, recover files or repair a configuration mistake. If Linux still boots, first use the VNC console and connectivity checks. You may be able to fix the problem without rescue media.
Do not select an installer, format a partition or reinstall the operating system when you want to keep its data. Encrypted disks need their recovery key. If the disk reports read errors or holds the only copy of important data, avoid write operations until you have a recovery plan.

​
Boot into rescue mode

1

Open ISO Boot

Sign in to the EDBB Control Panel, open your VPS and select Installation → ISO Boot. Save work and shut down the installed OS cleanly first if it still responds.
2

Mount the rescue ISO

Select the available Linux rescue image, such as SystemRescue, enable Force reset & boot from the selected ISO, then click Mount ISO. This interrupts the running VPS.
Installation and ISO Boot tabs with SystemRescue selected and Force reset checked

Select rescue media rather than an operating-system installer. Mounting with Force reset interrupts the VPS.

3

Open the console

Open the VNC console and follow the rescue image’s boot prompts. The shell now belongs to the rescue environment, not your installed Linux system.
SystemRescue console showing automatic root login and the rescue shell prompt

A successful rescue boot reaches the SystemRescue shell. This is separate from your installed Linux system.

​
Identify the installed system

List disks and filesystems before mounting anything:
Look for the VPS’s root filesystem. It may be a partition such as /dev/vda1 or a logical volume under /dev/mapper/. For LVM, inspect pvs, vgs and lvs; if the volumes are inactive, vgchange -ay activates them without formatting them. The following example assumes the root filesystem is ext4 on /dev/vda1. Replace the device with the one you identified. Mount it without journal replay for an initial read-only inspection:
For XFS, use mount -o ro,norecovery /dev/YOUR_ROOT_DEVICE /mnt/system instead. These options are filesystem-specific. Stop if you cannot identify the filesystem or mounting reports damage.

​
Inspect the problem

The installed system’s files are now under /mnt/system: You can copy needed files to a separate system before attempting repairs. Keep configuration backups and private keys secure.

​
Repair a known configuration mistake

Only proceed when you have identified the correct filesystem, have a backup and know which change to undo. Unmount the read-only filesystem before mounting it normally for editing. For the ext4 example:
If it cannot mount normally, stop; do not force it or run repair commands on a mounted filesystem. Back up the affected configuration file, then restore its known working version or correct the specific mistake. Preserve its ownership and permissions. Changing /etc/ssh in the rescue system changes the rescue environment. To edit your installation, use /mnt/system/etc/ssh. Likewise, stopping a service in rescue mode does not stop or repair the service in the installed system.

​
Return to the installed system

Close shells or file browsers using the mount if it is busy. In Installation → ISO Boot, unmount the ISO and boot from the local disk. Watch startup in the console, then test a fresh SSH connection and the affected application. If you need support, send the VPS IP, the console error and a description of the last change to support@edbb.com. Do not send passwords or recovery keys.