Hacker News Reader

Comment by chriswarbo | Hacker News Reader

You could harden a Linux server by killing the init process. That would cause a kernel panic, and prevent new processes running (e.g. shells). Existing processes like Web servers would continue to function, so you could put this at the end of the boot script :)

View on HN

*Groans*

...OK, after having a bit of a poke around I would respectfully ask for a demonstration/citation of this insanity.

My own experience is that systems that panic get rather warm, which I blindly assumed was because the kernel had entered an infinite loop. This turns out to be the case: https://github.com/torvalds/linux/blob/cbe232ab07ab7a1c7743c...

There are references to interacting with kgdb, booting a crash kernel, and delaying before reboot above the highlighted spot in the previous link. This suggests that panicking is very much a "everything is torn to shreds and never comes back again" event.

Annoyingly I can't SIGKILL, -TERM, or -SEGV `init` easily (PID 1 is special-cased by the kernel), and don't remember exactly how to fire up my kernel initramfs test harness, so I can't properly test this easily at the moment. Am very curious though.

I was able to crash a systemd version 245 init on an ubuntu machine via gdb a few minutes ago as a test. It did max out the virtual CPUs on the ESXI host. And while it did not cause the machine to reboot, it seems to have taken down all the processes with it, as SSH and NGINX also crashed and froze the console as well...

Uh huh. And when the web server process crashes, how do you restart it?

Second server with a relay that controls the power strip for the first server.

How do you recover control of the server, if remote?

You can't; it's so secure that nobody can control it. Except via the reboot switch :)

Can’t shell the computer if I break the OS ;)

Wow. Does this have a name?

if not any I will give one: self-destructive immutable server pattern

Reply on news.ycombinator.com