Gentoo Archives: gentoo-security

From: Tobias Klausmann <klausman@××××××××××××.de>
To: gentoo-security@l.g.o
Subject: Re: [gentoo-security] iptables window of opportunity at startup
Date: Wed, 08 Feb 2006 11:29:13
Message-Id: 20060208112230.GA32287@eric.schwarzvogel.de
In Reply to: Re: [gentoo-security] iptables window of opportunity at startup by Francois Toussenel
1 Hi!
2
3 On Tue, 07 Feb 2006, Francois Toussenel wrote:
4
5 > On Sun, 5 Feb 2006 13:29:55 +0100 Tobias Klausmann <klausman@××××××××××××.de> wrote:
6 >
7 > > Which *should* make iptables start before net.* (maybe except
8 > > net.lo). And sure enough, the boot sequence is:
9 >
10 > This depends on the runlevels in which you have iptables and net.eth0.
11 > Could you please post the output of the following command?
12 >
13 > # rc-update show | grep 'iptables\|net\.'
14 >
15 > By having iptables in boot and net.eth0 in default, iptables starts
16 > before net.eth0, but it also stops before services and of course
17 > net.eth0. Does somebody know a setting to avoid that?
18
19 I'm using the defaults for both (i.e. I did what's in the install
20 handbook):
21
22 $ rc-update show | grep 'iptables\|net\.'
23 iptables | default
24 net.eth0 | default
25 net.lo | boot
26
27 I really don't understand what happened on the original poster's
28 machine. My (wild) guess is, that somehow parallel startip messed
29 it up, but that would be a bug in the parallel startip code.
30
31 > (I would add that one might want to never respond to pings, for
32 > instance, so starting iptables between net.eth0 and services seems not
33 > enough.)
34
35 Why (outside of s specific attack in that area) would one *not*
36 respond to pings? Outside from a specific attack in that area
37 happening, I see no reason to do so.
38
39 Regards,
40 Tobias
41 --
42 You don't need eyes to see, you need vision.
43 --
44 gentoo-security@g.o mailing list

Replies

Subject Author
Re: [gentoo-security] iptables window of opportunity at startup Oliver Schad <o.schad@×××.de>