There are times when we modify our kernels to optimize them, upgrade them, or tinker around with them. Unfortunately for Linux-based operating systems, this means that there is the possibility of kernel panics – the equivalent to a “blue screen of death” on Windows. It’s important that you know how to recover from these, as any KVM VPS (including Vultr) or dedicated server that you use has its own kernel.
For this article, I’ll explain how to recover from a kernel panic on a Vultr VPS. I will be using an ISO from the SystemRescueCd project.
Step 1: Finding the error
From the Vultr control panel, select your VPS and hit VNC console. Attempt to boot the server, and you’ll be able to see the error. Take note of this so that you can repair it.
Step 2: Mounting a custom ISO from the control panel
Click the “ISO” option, and input the SystemRescueCd ISO URL. You may visit the official site for newer images.
From there, head back to your VPS controls and click “Mount ISO”.
You will be greeted with a pop-up screen by pressing the VNC console. Click “Enter” on the first option, and your system will start up.
Step 3: Rescuing your kernel
Now that you have access to a shell prompt, create a directory to mount your VPS’s disk.
Now, mount the disk.
mount /dev/vda1 /rescuedisk
You now have access to your files. At this point, you can choose whether to copy your files to a remote server, or research the kernel panic’s message that you took note of earlier.
An example of a kernel panic issue would be a missing
/etc/shadow file, or any missing system file. You could replace a missing file with a backup by copying over
/etc/shadow. For example:
cd /rescuedisk cp etc/shadow- etc/shadow
Once you have finished fixing the issue, unmount the ISO from the Vultr control panel and reboot your VPS by typing
reboot in the prompt. If your issue was fixed, then your VPS will boot normally.