lkml.org 
[lkml]   [1996]   [Jun]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: boot parameters, init= and security issues
Date
Hi,

> >> Generally, though I am in favor of your proposal. Providing as many
> >> boot-time security measures as possible sounds like a *very* good idea
> >> --- but maybe I am just a little bit paranoid :-)

this reminds me of root NFS which tries to fallback to a floppy boot when
it doesn't get a RARP/BOOTP answer after a certain time. I think this
feature is a bad idea in terms of security:

- let a diskless machine boot its kernel
- unplug the network cable before the kernel tries to get RARP/BOOTP answers
- wait the the kernel gives up waiting for a RARP/BOOTP answer
- insert your favourite bootdisk

You may say given physical access to a machine this class of attacks is always
easy. Well, you're right. It's just that the NFS root stuff just makes it
*too* easy. There is a big difference between just removing and reattaching a
network cable (no tools required) and other, more conspicuous, manipulations.

Aside of this this fallback action when no RARP/BOOTP answer is received is
in my experience never a helpful action; in case of a large network like my
university's it's even the wrong thing after a complete power failure. (The
right thing would be to wait for the servers coming back online. Given the
problems with our larger servers this might mean waiting for hours.)

Ralf


\
 
 \ /
  Last update: 2005-03-22 13:37    [W:0.026 / U:6.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site