Firewall.txt v2.0

Henning Rohde Rohde.Henning at
Mon Sep 3 06:29:31 PDT 2001

Hi Ian,

just some few minutes after my mail a new version of iptables was released!

Please apply the attached version, not the text I sent you last time.

Sorry for inconvenience,


On Sun, 2 Sep 2001, Henning Rohde wrote:

> Date: Sun, 2 Sep 2001 19:03:33 +0200 (CEST)
> From: Henning Rohde <hr at>
> To: hints at
> Subject: Firewall.txt v2.0
> Hi Ian,
> although I promised that the last revision would be the last before
> BLFS for the firewalling-hint, I decided that the help of Jeff Baumann
> should be submitted for itself.
> Please put the attached file as version 2.0 onto CVS.
> Changes: intensive improvements on wording
> 	 rewritten chapter C
> 	 further minor changes
> Thank you and have a nice sunday,
> 	Henning
-------------- next part --------------
TITLE:		Firewall.txt
		Kernel   >= 2.4.9
		iptables >= 1.2.3
AUTHOR:		Henning Rohde	Henning.Rohde(at)

	Question:	What's a firewall?
 	Answer:		a fire-resistant wall separating a building into 
			departments, designed to prevent the spread of fire
And on Networks?	A box that restricts the malicious (eg crackers, 
			worms, trojans) out of your intranet.
How Do I build one?	fetch iptables and read the following:

Personally I prefer the following definition:
Answer:		A wall of fire that only the Saints can pass through!
@Networks:	Just a box, that permits only sane packets to pass!

The general purpose of a firewall is to save the labour of securing every
box by securing a single firewall.
This means that the firewall is a single point of failure, but it makes
the admin's life a lot easier.

But please, don't assume that having a firewall makes careful 
configuration redundant!

If you knew every daemon or service on every machine was correctly configured
and trustworthy, and trusted every user accessing your services to cause 
no harm, you wouldn't need to do firewalling!
But if you'd like to choose which services are accessible only from your 
intranet, which machines or applications are allowed to have internet access,
or if you don't trust some of your apps or users, you might benefit by using
a firewall.

When you read the word "firewall" there's more than one way to interpret it:
a) "Personal Firewall":	
	Program sold by e.g., Symantec, that is claimed to secure a home / 
	desktop-pc with internet access. Quite relevant for users being 
	always online with (flat rate) broadband links.
b) Firewall as it was originally meant: 
	A box placed between the internet and intranet doing, besides routing, 
	nothing but protecting the intranet. 
	This could include the function of masquerading: rewriting IP-headers
	of the packets it routes from clients with private IP-adresses onto 
	the internet, so that they seem to come from the firewall itself.
c) Firewall offering services:
	Some old box you may have retired and nearly forgotten, doing (B), 
	but offering a bunch of services, e.g., web-cache, mail, etc.
	This may be very commonly used for home networks, but it severely 
	violates some principles of (B).
d) Firewall with a demilitarized zone	[not described here]:
	Doing (B), but giving public access to some branch of your network,
	that is, because of public IP's and a physically separated structure,
	neither considered to be part of the inter- nor intranet.
	Here those servers are connected that must be easily accessible
	from both the inter- and intranet. The firewall protects them all.
e) Packetfilter / partly accessible net	[only partly described here, see (C)]:
	Doing (B) but permitting only selected services to be accessible,
	sometimes only by selected internal users or boxes; mostly used in
	highly secure business contexts, sometimes by distrusting employers.
	This was the common configuration of a firewall at the time of 
	the Linux 2.2 kernel.
	It's still possible to configure a firewall this way, but it makes 
	the rules quite complex and lengthy.

This document is meant as an introduction to how to setup a firewall. 
I am not, nor do I pretend to be, a security expert.		;-)
I am just some guy who still has not read enough and whose computers
still like to play tricks on him if he wants to tweak them.	;-)
Please, I am writing this to help people get acquainted with this subject, 
and I am not ready to stake my life on the accuracy of what is in here.
(Taken from, slightly modified)

The scripts quoted here are not at the very least meant to give you 
the one and perfect firewall, being perfect until the universe collapses.
They may not fit into any imaginable configuration and may not prevent 
any imaginable attack.

The purpose of this text is simply to give you a hint on how to get started
with firewalling.

Customization of these scripts for your specific situation is likely to be 
necessary for an optimal configuration, but you should make a serious study 
of the iptables documentation and firewalling in general before hacking away.
Have a look at Appendix I. There you'll find a list of URLs that contain quite 
comprehensive information about building your own firewall.

|Getting a firewalling-enabled Kernel|
If you want your Linux-Box to do firewalling you must first ensure that you 
have an appropriate kernel and the appropriate tools:

But, before you do a 'make menuconfig', consider patching your kernel with 
the latest iptables enhancements. To get them, download the latest version 
of iptables from

Having current kernel sources in /usr/src/linux, unpack iptables and 'cd' into 
its top-level directory.
Now enter 'make pending-patches', as an user who is allowed to patch the kernel.
There's no need to apply every patch, but e.g. the ipt_REJECT-checkentry-patch
could be sensible, depending on your needs. 

Applying a patch may result in errors, mostly because the hooks for the
patches have changed or because the runme script doesn't recognize that 
a patch has already been incorporated.
If you do get errors don't worry too much; the important thing is that
iptables has scanned the kernel for the already accepted / incorporated 
patches. The default kernel should be adaequate for most needs!

Now configure the kernel:
Personally I prefer to have a maximally modularized kernel, but for highest
security you could configure the kernel with this code built in:
    Networking options:
	Network packet filtering	= CONFIG_NETFILTER
	IP: TCP/IP networking		= CONFIG_INET
	IP: advanced router		= CONFIG_IP_ADVANCED_ROUTER
	IP: verbose route monitoring	= CONFIG_IP_ROUTE_VERBOSE
	IP: TCP Explicit Congestion Notification support
	IP: TCP syncookie support	= CONFIG_SYN_COOKIES
	IP: Netfilter Configuration: every option
					= CONFIG_IP_NF_*
		BUT NO ipchains- and ipfwadm-compatibility.
    Although you are unlikely to have enabled it, please make sure not to 
    select "Fast switching"		w\ CONFIG_NET_FASTROUTE
    because it would currently bypass your firewalling-rules.

Now compile and install your new kernel, update your bootloader and reboot.

|Building iptables|
Before compiling you might want to edit the Makefile to adapt install-dir's.
Now compile and install iptables via 'make && make install'.

|Now we can start to build your Firewall|
A Personal Firewall is supposed to let you access the all services offered 
on the internet, but keep your box secure and your data private.

Below is a slightly modified version of Rusty Russel's recommendation,

# Insert connection-tracking modules	(not needed if built into kernel).
modprobe ip_tables
modprobe iptable_filter
modprobe ip_conntrack
modprobe ip_conntrack_ftp
modprobe ipt_state
modprobe ipt_LOG
# free output on any interface to any ip for any service   (equal to -P ACCEPT)
iptables -A OUTPUT					  	-j ACCEPT
# permit answers on already established connections
# and permit new connections related to established ones (eg active-ftp)
iptables -A INPUT	-m state --state ESTABLISHED,RELATED	-j ACCEPT
# Log everything else:	What's the latest exploitable vulnerability?
iptables -A INPUT		-j LOG --log-prefix "FIREWALL:INPUT  "
# set a sane policy:	everything not accepted > /dev/null
iptables -P INPUT		DROP
iptables -P FORWARD		DROP
iptables -P OUTPUT		DROP
# be verbose on dynamic ip-adresses	(not needed in case of static IP)
echo 2 > /proc/sys/net/ipv4/ip_dynaddr
# disable ExplicitCongestionNotification - too many routers are still ignorant
echo 0 > /proc/sys/net/ipv4/tcp_ecn

His script is quite simple, but simply surfing the internet you are unlikely 
to exceed its limits.

Even if you have daemons / services running on your box, these should be
inaccessible everywhere but on your box itself.
The case to be cautious about is misconfigured daemons that could broadcast
to the public to announce their service.
If you have such daemons (e.g., cups, samba) running and if you are not 
confident in their configuration, restrict OUTPUT and INPUT, see (C) and (E).

A true Firewall has two interfaces, one connected to an intranet, 
in this example, eth0, and one connected to the internet, here, ppp0.
To provide the maximum security against the box itself being broken into, e.g.,
by exploiting an offered service, make sure that there are no servers running 
on it, especially not X11 et al., and, as a principle, that it does not itself
access any services:

echo -n	"You're using the example-config "
echo	"for a setup of a firewall "
echo -n	"from the firewalling-hint "
echo	"written for LinuxFromScratch. "
echo -n	"This example is far from being "
echo	"complete, it is only meant "
echo	"to be a reference. "
echo -n	"Firewall security one could rely "
echo	"on is a complex issue, "
echo -n "that exceeds the scope of the "
echo	"quoted configuration rules. "
echo -n "You can find some quite "
echo	"comprehensive information "
echo	"about firewalling at Appendix I of "
echo -n	""
echo	"."
echo	"Be cautious!"
modprobe ip_tables
modprobe iptable_filter
modprobe ip_conntrack
modprobe ip_conntrack_ftp
modprobe ipt_state
modprobe iptable_nat
modprobe ip_nat_ftp
modprobe ipt_MASQUERADE
modprobe ipt_LOG
modprobe ipt_REJECT		# needed for (C), example 4
# allow local-only connections
iptables -A INPUT	-i lo					-j ACCEPT
iptables -A OUTPUT		-o lo				-j ACCEPT
# allow forwarding
iptables -A FORWARD -m state --state ESTABLISHED,RELATED 	-j ACCEPT
iptables -A FORWARD -m state --state NEW	-i ! ppp+	-j ACCEPT
# do masquerading    (not needed if intranet is not using private ip-adresses)
iptables -t nat -A POSTROUTING  -o ppp+				-j MASQUERADE
# Log everything for debugging (last of all rules, but before DROP/REJECT)
iptables -A INPUT			-j LOG --log-prefix "FIREWALL:INPUT  "
iptables -A FORWARD			-j LOG --log-prefix "FIREWALL:FORWARD"
iptables -A OUTPUT			-j LOG --log-prefix "FIREWALL:OUTPUT "
# set a sane policy
iptables -P INPUT		DROP
iptables -P FORWARD		DROP
iptables -P OUTPUT		DROP
# be verbose on dynamic ip-adresses (not needed in case of static IP)
echo 2 > /proc/sys/net/ipv4/ip_dynaddr
# disable ExplicitCongestionNotification
echo 0 > /proc/sys/net/ipv4/tcp_ecn
# activate TCPsyncookies
echo 1 > /proc/sys/net/ipv4/tcp_syncookies
# activate Route-Verification = IP-Spoofing_protection
for f in /proc/sys/net/ipv4/conf/*/rp_filter; do
	echo 1 > $f
# activate IP-Forwarding 
echo 1 > /proc/sys/net/ipv4/ip_forward

With this script your net should be sufficiently secure against external 
attacks: your intranet should be invisible because it's masqueraded or, 
at least, no one should be able to setup a new connection to its services, 
and your firewall should be immune because there are no services running 
that a cracker could attack.

If you are in the need of stronger security (e.g., against DOS, connection 
highjacking, spoofing, etc.) see Appendix.1 and start to read a bit!

This scenario is not too different from (B), but you'd like to offer some 
services to your intranet. This gets relevant when you want to admin your 
box from another host on your intranet or use it as a proxy or a nameserver.

Be cautious: every service you do offer and have enabled makes your setup 
more complex and your box less secure. See introduction to (B)!

If the services you'd like to offer do not need to access the internet 
themselves, like an internal-only mail- or name-server, it's quite simple 
and should still be acceptable from a security standpoint.
Just add the following lines _before_ the logging-rules of script (B):
iptables -A INPUT	-i ! ppp+				-j ACCEPT
iptables -A OUTPUT	-o ! ppp+			  	-j ACCEPT

If your daemons have to access the web themselves, like squid would need to, 
you could open OUTPUT generally and restrict INPUT.
iptables -A INPUT	-m state --state ESTABLISHED,RELATED	-j ACCEPT
iptables -A OUTPUT					  	-j ACCEPT

Personally I am not so confident in the configuration of my own daemons that 
I would feel comportable with this, I think that this way it is too risky. 
Every misconfigured daemon and any trojan could easily open a connection to 
the public and expose private data.
As a result, I prefer to restrict not only INPUT but both INPUT and OUTPUT 
on all ports except those that it's absolutely necessary to have open.
Which ports you have to open depends on your needs: you will find them 
by looking at the FIREWALL-lines in your log-files.

Have a look at the following examples:
1) Squid is requesting data:
iptables -A OUTPUT -p tcp --dport 80				  -j ACCEPT
iptables -A INPUT  -p tcp --sport 80 -m state --state ESTABLISHED -j ACCEPT

2) Your caching-nameserver (e.g., dnscache) does its lookups via udp:
iptables -A OUTPUT -p udp --dport 53				  -j ACCEPT
iptables -A INPUT  -p udp --sport 53 -m state --state ESTABLISHED -j ACCEPT

3) Alternativly, if you want to ping your box to ensure it's still alive:
iptables -A INPUT  -p icmp -m icmp --icmp-type echo-request	-j ACCEPT
iptables -A OUTPUT -p icmp -m icmp --icmp-type echo-reply	-j ACCEPT

These are only examples to show you some of the capabilities of the new 
firewalling-code in Linux-Kernel 2.4. Have a look at the man-page of iptables.
There you will find more of them. The port-numbers you'll need for this can be
found in /etc/services, if you didn't find them in your logfile.

If you add any of your offered or accessed services such as the above,
maybe even in FORWARD and for intranet-communication, and delete the general 
clauses, you get an old fashioned packet filter, not unlike that one 
mentioned in (E).

4) If you are frequently accessing ftp-servers or enjoy chatting you might 
notice a certain delay because some oldfashioned implementations of these 
daemons insist on querying an identd on your box for your username.
Although there's no harm in this, I wouldn't recommend using an
identd, as many of them are known to be vulnerable.
To avoid these delays you could reject the requests with a 'tcp-reset':
iptables -A INPUT  -p tcp --dport 113	-j REJECT --reject-with tcp-reset
iptables -A OUTPUT -p tcp --sport 113	-m state --state RELATED   -j ACCEPT

5) To log and drop invalid packets, mostly harmless packets that came in 
after netfilter's timeout, sometimes scans.
iptables -I INPUT 1	-p tcp	-m state --state INVALID	-j LOG	\
	--log-prefix "FIREWALL:INVALID"
iptables -I INPUT 2	-p tcp	-m state --state INVALID	-j DROP

6) Anything coming from the outside should not have a private address, 
this is a common attack called IP-spoofing:
iptables -t nat -A PREROUTING	-i ppp+	-s	-j DROP
iptables -t nat -A PREROUTING	-i ppp+	-s		-j DROP
iptables -t nat -A PREROUTING	-i ppp+	-s	-j DROP

7) To simplify debugging and be fair to anyone who'd like to access a service
you have disabled, purposely or by mistake, you should REJECT those packets 
that are dropped.
Obviously this must be done directly after logging as the very last lines 
before the packets are dropped by policy:
iptables -A INPUT	-i eth+		-j REJECT
iptables -A OUTPUT		-p icmp --icmp-type 3		-j ACCEPT

Finally, I'd like to remind you of one fact we must not forget:
The effort spent attacking a system corresponds to the value the cracker
expects to gain from it.
If you are responsible for such valuable assets that you expect great
effort to be made by potential crackers, you hopefully won't be in need 
of this hint!

Be cautious!
		Henning Rohde
	(Henning.Rohde at

PS: And always do remember:
	SecureIT is not a matter of a status-quo 
	but one of never stopping to take care!
PPS: If any of these scripts fail, please tell me. I will try to trace
     any faults.

Jeff Bauman	jbauman(at)
	He calls himself to be paranoid, at least in matters of security;  ;-)
	I'd like to thank him for our discussion about security-related 
	philosophies and furtherly for reviewing the hint and for giving
	not only a bit of help with the fine points of wording.

Nowadays, we must face the threat of denial of service attacks (DoS) even 
against private users (this seems to be quite common if you do online-gaming),
trojans (read on IRC for commands), and worms exploiting the internet as if
someone was doing a blitzkrieg.

There may be ways to protect both your router and your intranet, but any
solution I'm able to give here could become insufficent tomorrow
and would give you a false sense of security.
If you are really concerned, this is not the document to help you out!

But have a look, here's where I'd suggest you start reading: (IIRC outdated!) +s-fire2.html (German & outdated, but very comprehensive)

If a link proves to be dead or if you think I missed one, please mail!

If you need to turn firewalling off, this script will do it:

# deactivate IP-Forwarding 
echo 0 > /proc/sys/net/ipv4/ip_forward
iptables -Z
iptables -F
iptables -t nat         -F PREROUTING
iptables -t nat         -F OUTPUT
iptables -t nat         -F POSTROUTING
iptables -t mangle      -F PREROUTING
iptables -t mangle      -F OUTPUT
iptables -X
iptables -P INPUT       ACCEPT
iptables -P FORWARD     ACCEPT
iptables -P OUTPUT      ACCEPT

If you'd like to have a look at the chains your firewall consists of and 
the order in which the rules take effect:

echo "iptables.mangling:"
iptables -t mangle	-v -L -n --line-numbers
echo "iptables.nat:"
iptables -t nat		-v -L -n --line-numbers
echo "iptables.filter:"
iptables		-v -L -n --line-numbers

More information about the hints mailing list