Table of Contents
These files implement a Linux netfilter target that changes the IPv6 address of packets. The address change is done checksum neutral, thus no checksum re-calculation for the packet is necessary. You can change the IPv6 source address of outgoing packets as well as the IPv6 destination address of incoming packets. This allows you to map an internal IPv6 address range to a second, externally used IPv6 address range. IPv6 address mapping is not very similar to IPv4 network address translation, but one can describe it as some sort of stateless NAT. The implementation is based on the expired IETF discussion paper published here:
http://tools.ietf.org/html/draft-mrw-behave-nat66-02
Using MAP66 rules together with connection tracking rules such as
--ctstate is currently untested and may not work or
may cause dysfunctions.
MAP66 implements two pieces of software: a shared library that extends the ip6tables command and a Linux kernel module. The shared library file adds the '-j MAP66' target to the ip6tables command. To build and install, you need ip6tables installed as well as the necessary headers. The Linux kernel module requires the Linux source file tree and kernel configuration files to compile. On a Debian / Ubuntu, the following command prepares the build environment:
sudo apt-get install build-essential linux-headers iptables-dev
Unpack the source tgz archive below /usr/src,
change to the new sub-directory and issue "make" to build. If this
compiles without errors, install the ip6tables extension with the
following command:
sudo make install
The kernel module (ip6t_MAP66.ko for
Linux-2.6 or ip6t_MAP66.o for Linux-2.4) is not
automatically installed nor loaded into the kernel. You can copy the
kernel module file manually, e.g. with sudo cp ip6t_MAP66.ko
/lib/modules/$(uname -r)/.
If the next system update needs to install a new kernel version, you
also need to re-compile/re-install the MAP66 kernel module. With Debian /
Ubuntu, this can be automated with the Dynamic Kernel Module Support
Framework (DKMS). For this, the dkms.conf file is
included with the MAP66 source file package. Install DKMS with the
following command:
sudo apt-get install dkms
If not already in place, move/unpack the MAP66 source file archive
below /usr/src/. To register the MAP66 source to DKMS
and compile/install, issue these commands:
sudo dkms add -m ip6t_MAP66 -v 0.6 sudo dkms build -m ip6t_MAP66 -v 0.6 sudo dkms install -m ip6t_MAP66 -v 0.6
Read DKMS details here: https://wiki.kubuntu.org/Kernel/Dev/DKMSPackaging. Also, there's a pre-packed binary with DKMS for Debian / Ubuntu hosted on my packet repository: http://sven-ola.dyndns.org/repo/.
You always need to add two ip6tables-rules to your netfilter configuration. One rule matches outgoing packets and changes their IPv6 source address. The second rule matches incoming packets and reverts the address change by altering their IPv6 destination address. To following commands correspond to the “Address Mapping Example” given in the IETF discussion paper:
ip6tables -t mangle -I POSTROUTING -o eth0 -s FD01:0203:0405::/48 -j MAP66 --src-to 2001:0DB8:0001::/48 ip6tables -t mangle -I PREROUTING -i eth0 -d 2001:0DB8:0001::/48 -j MAP66 --dst-to FD01:0203:0405::/48
This example is also printed to the screen if you issue
ip6tables -j MAP66 --help. By design, you cannot
use an arbitrary prefix length. Only /112, /96 .. /16 are supported,
because the MAP66 kernel module checksum re-calculation is made with
16-bit integers.
For each packet, the MAP66 kernel module also compares the
packet's source address to all IPv6 addresses assigned to the outgoing
interface. If a match is found, the packet's source address is not
mapped. The same comparison happens on the incoming packet's destination
address. The comparison requires some CPU resources, especially if the
interface has a large number of assigned IPv6 addresses. If you are sure
that the mapping cannot match the IPv6 address of the interface (e.g.
the mapping rule defines a mapping prefix that cannot result in the
interface address) you can switch off the comparison. Add the
--nocheck parameter to the ip6tables command
for this.
The following explanation details a living example from the wireless mesh network that is mentioned under Motivation (see below). Throughout the mesh network, a private IP address range is used. The ULA prefix is fdca:ffee:babe::/64. All mesh nodes derive their IPv6 interface addresses by correlating the ULA prefix with the EUI48 (“MAC address”) of the respective network adapter.
There is a Debian based virtual machine that should act as one
IPv6 Internet gateway for the mesh. You can reach the virtual machine's
web service via IPv4 under http://bbb-vpn.freifunk.net.
To experiment with IPv6, a SIXXS static tunnel setup has been
added. The following /etc/network/interfaces file
provides the configuration for IPv6:
auto sixxs
iface sixxs inet6 v4tunnel
address 2001:4dd0:ff00:2ee::2
netmask 64
local 77.87.48.7
endpoint 78.35.24.124
ttl 64
up ip link set mtu 1280 dev $IFACE
up ip route add default via 2001:4dd0:ff00:2ee::1 dev $IFACE
up ip addr add 2001:4dd0:fe77:ffff::1/48 dev $IFACEAs you can see, the virtual machine has an IPv6 prefix of
2001:4dd0:fe77::/48. The /48 prefix length includes 65536 /64 subnets
where this setup occupies the “ffff” subnet. The machine is
therefore reachable via http://[2001:4dd0:fe77:ffff:1]/.
The netfilter setup of this machine is initialized with some basic
precautions. The “-j ACCEPT” rules prevents any further
matching on the POSTROUTING chain for local network packets. Most
notably, we also block packets that leave the machine on the receiving
interface so we cannot participate as a “reflector” in DDOS
with the --dst-swap option used below.
ip6tables -t mangle -A FORWARD -p tcp --tcp-flags SYN,RST SYN -j TCPMSS --clamp-mss-to-pmtu ip6tables -t mangle -A POSTROUTING -o sixxs -s 2001:4dd0:fe77::/48 -j ACCEPT ip6tables -t mangle -A POSTROUTING -o sixxs -s 2001:4dd0:ff00:2ee::/64 -j ACCEPT ip6tables -t filter -A FORWARD -i sixxs -o sixxs -j DROP
You may also stumble over the MSS-clamping rule. While IPv6
defines, that path MTU detection via ICMPv6 must be supported by any
host, sometimes path MTU detection does not work. The SIXXS tunnel
uses an MTU of 1280 byte. To get the following command working on my
PC, I needed to add the above MSS-clamping rule on the gateway:
wget --prefer-family=IPv6 -O -
http://6to4.nro.net/.
The netfilter setup then includes the following command sequence to realize mapping from the private fdca:ffee:babe::/64 prefix to the globally valid IPv6 addresses and vice versa.
ip6tables -t mangle -A POSTROUTING -o sixxs -s fdca:ffee:babe::/64 \
-j MAP66 --src-to 2001:4dd0:fe77:100::/64 --nocheck
ip6tables -t mangle -A PREROUTING -i sixxs -d 2001:4dd0:fe77:100::/64 \
-j MAP66 --dst-to fdca:ffee:babe::/64 --nocheckBecause for these IPv6 networks the external prefix length (/64)
is smaller than the internal prefix length (/48), we are sure that
mapped addresses cannot match the interface address. For example:
2001:4dd0:fe77:100::/64 cannot be mapped / converted to
2001:4dd0:fe77:ffff::1 in this context. For this reason, we can use the
--nocheck speedup here.
As an extension, the --csum option
updates the packet's checksum instead of updating some of the IP address
bits. This only works for some of the transported protocols (TCP, UDP,
ICMP6, DCCP) because each protocol uses another offset for storing the
checksum.
This option can be used to map exactly one IPv6 address to another, hence using --src-to or --dst-to with /128.
In our mesh network, we have routers with old and new firmwares. While the old firmware uses a single /64 ULA prefix for all nodes and interfaces, the new firmware uses an arbitrary /64 prefix derived from the default route address. The new firmware always use ::1 for the host bits, e.g. 2002:abcd:efgh:xxxx::1/64 if the neighbour's default route is via 6to4, or fdca:ffee:babe:xxxx::1/64 if the neighbour's default route is via ULA. Anyhow, those routers “waste” 64 bits of address space because of the fixed host bits.
To map those arbitrary prefixes to the prefix used on the Internet
gateway, we therefore can swap net bits with host bits and map the
result to the gateway's prefix. Example: if a mesh node uses
2002:1234:5678:9abc::1/64 we can map this to
2001:4dd0:fe77:1:2002:1234:5678:9abc/64 and send to / receive from the
Internet. For this, MAP66 support --src-swap and
--dst-swap options. Here's the working
example:
ip6tables -t mangle -A POSTROUTING -o sixxs \
-m u32 --u32 "0x08 & 0xe0000000 = 0x20000000 && 0x10 = 0 && 0x14 & 0xffffff00 = 0" \
-j MAP66 --nocheck --src-swap --src-to 2001:4dd0:fe77::/48
ip6tables -t mangle -A POSTROUTING -o sixxs \
-m u32 --u32 "0x08 & 0xfe000000 = 0xfc000000 && 0x10 = 0 && 0x14 & 0xffffff00 = 0" \
-j MAP66 --nocheck --src-swap --src-to 2001:4dd0:fe77::/48
ip6tables -t mangle -A PREROUTING -i sixxs \
-m u32 --u32 "0x18 = 0x20014dd0 && 0x1c & 0xffffff00 = 0xfe770000 && 0x20 & 0xe0000000 = 0x20000000" \
-j MAP66 --dst-swap --dst-to 0::/48 --nocheck
ip6tables -t mangle -A PREROUTING -i sixxs \
-m u32 --u32 "0x18 = 0x20014dd0 && 0x1c & 0xffffff00 = 0xfe770000 && 0x20 & 0xfe000000 = 0xfc000000" \
-j MAP66 --dst-swap --dst-to 0::/48 --nocheckThe above example uses the xtables u32 match to filter out packets with IPv6 addresses that uses host bits containing ULA (fc00::/7) or IANA Internet (2000::/3).
Swapping / rotating 16 bit integers within IP packet headers is
checksum neutral by design. Also, swapping / rotating 8 bit integers
can be compensated with the following formula: addr[w] =
add16(addr[w], add16(oldpartialcsum,
~byteswap(oldpartialcsum))).
With Ubuntu and eventually with RedHat, you will notice that your
browser does not show the IPv6 version of a web site that is multi-homed
when using ULA addresses for your IPv6 Internet connection. The reason for
this is an add on to the RFC 3484 rules that is compiled into the Ubuntu
libc. The pre-installed /etc/gai.conf file will give
you a hint on this.
In short: the getaddrinfo() library function rates a private IPv4
address higher than the ULA IPv6 address when choosing the transport
protocol for a new Internet connection if this add on to the RFC 3484
rules is compiled in. For this reason, you may want to change the
precedence rules within /etc/gai.conf (see Change gai.conf) or use another
prefix (see Use Changed Internal Address).
The getaddrinfo() library function manages lists of label,
precedence, and scope4 type entries. If the
/etc/gai.conf file does not provide a single entry
for a particular type, the compiled-in list is used. For this reason,
you cannot uncomment a single entry to overwrite the default. You need
to uncomment all entries of a particular type for this. The
“label” lines compare source addresses, the
“precedence” lines compare destination addresses.
Procedure 1. Change IPv6 Precedence
Open the /etc/gai.conf file as root user,
e.g. by executing sudo nano
/etc/gai.conf.
Remove the leading hash character from the 8 lines starting with “#label”.
Re-add the hash character to the line stating “#label fc00::/7 6”.
Save the file.
Restart your browser and re-try to browse to a multi-homed web site.
The above procedure removes the difference between standard IPv6 source addresses and ULA type private IPv6 source addresses. Anything else is unchanged.
As an alternative solution, you may use an arbitrary address
prefix in your LAN that is not mentioned in the
gai.conf file nor compiled in. This will work but
introduces a double mapping: one map (Inet-ULA) on the Internet gateway
router and a second map (ULA-Intern) on the internal router.
While the well known IPv4 addresses 10.0.0.0/8, 172.16.0.0/12, and 192.168.0.0/16 still exist, it is unlikely that their 6to4 counterparts 2002:0a00::/24, 2002:ac10::/28, and 2002:c0a8::/32 will be routed on the Internet. Sadly, the Ubuntu defaults penalize 6to4 addresses also.
If you already deployed ULA addresses in your network, you may be interested in a solution that runs on my Freifunk router. The router uses the IPv6 prefix that is reserved for documentation purposes on it's LAN interface. Within the OLSR-based mesh network, any interface uses an fdca:ffee:babe::/64 prefix. The following internal mapping is configured for this:
ip6tables -t mangle -I PREROUTING -i br0 -s 2001:0DB8::/64 -j MAP66 --src-to fdca:ffee:babe::/64 --csum ip6tables -t mangle -I POSTROUTING -o br0 -d fdca:ffee:babe::/64 -j MAP66 --dst-to 2001:0DB8::/64 --csum
To prevent the mapped packets to vanish via the default route and
to overcome mac address lookups during the routing process, I also added
these prefixes to the router's /etc/radvd.conf as
well as (host) routes pointing to the “br0” interface for
both prefixes.
My Internet access at home is realized by a wireless community mesh network not owned by me. The mesh is operated with small embedded devices (nodes aka. WLAN routers) that are interconnected via radio links (WLAN IBSS / AdHoc). Routing is done with a specialized protocol such as Batman or OLSR. The routing protocol selects the nearest out of a dozen Internet gateways and configures a default route or an IPIP tunnel accordingly. Each Internet gateway is connected to a different ISP and provides the service with the help of IPv4 network address translation (NAT). Using NAT has the following effects:
Address amplification - something not necessary with IPv6 any more
Anonymization - nice to have as an option but not mission critical
ISP independence - no reverse routing, no "buy-a-number-range"
The last point is mission critical. One can obtain a provider independent IPv6 address range, but you need the cooperation of an ISP to use that address range for Internet connectivity. If you e.g. move to another ISP you need your address range to be re-routed to your new location.
ISP independence is also possible with some tunneling technique, such as VPN or mobile IP. Tunneling can be implemented on client PCs and Internet gateways/servers one day. But there is no need to implement the same tunneling technique on every mesh node. Why? Because the mesh nodes can use private IP addresses (or "ULA") to transport the tunnel data between the client PC and the gateway/server. Each tunneling technique typically needs a single instance (the "server") which forms a single point of failure. Rule-of-thumb1: avoid a SPOF for the infrastructure. Rule-of-thumb2: KISS (keep it simple stupid).
Using private IP addresses on the mesh nodes has a drawback: mesh node software updates e.g. a download via HTTP from an Internet server is not possible. This is where I start to think: “hey, some kind of address mapping may be nice to have”. While opening Pandora's NAT66 box, I discovered that IPv6 nerds do not like the acronym. It is always a good tactic in info wars to rename, hence the name "MAP66".
// Sven-Ola