Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Jun 2004 13:04:11 -0500
From:      Craig Boston <craig@meoqu.gank.org>
To:        freebsd-stable@freebsd.org, Skylar Thompson <skylar@cs.earlham.edu>
Subject:   Re: Maximum uptime 497 days?
Message-ID:  <200406301304.11783.craig@meoqu.gank.org>
In-Reply-To: <20040630175241.GC54215@quark.cs.earlham.edu>
References:  <200406281706.11188.matt@fruitsalad.org> <200406281644.i5SGiM0h097809@lurza.secnetix.de> <20040630175241.GC54215@quark.cs.earlham.edu>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday 30 June 2004 12:52 pm, Skylar Thompson wrote:
> A lot of security holes can be patched without rebooting. In general, only
> kernel updates strictly require a reboot. There have been a few kernel
> security vulnerabilities released in the past couple years, but a lot of
> them are for DoS attacks, not privelege escelation.

Also, _in theory_, even many kernel bugs can be patched without rebooting.
A kernel module can bypass an affected function, for example by replacing 
syscall table entries.  It takes a lot of work and knowledge of the guts of 
the kernel, but it is possible.

I've never done this myself but have seen it used in environments such as 
massive virtual hosting (we may be talking about hundreds of potentially 
affected servers, each with dozens of users), where a reboot is costly and 
painful.

Craig



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200406301304.11783.craig>