Skip to contentSkip to conversation

Trivial IPv6-for-Fedora HOWTO

DISCUSSION Jeff Garzik 7 participants 9 messages Aug 22, 2004 – Aug 23, 2004

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

So, thanks to David Woodhouse for showing me how to do this. IPv6 appears to be very, very close to a Just Works(tm) state.

These instructions are for Fedora Core 2 users, and describe how to set up IPv6 automatically tunnelling (6to4) on an IPv4 network. If you are stuck on an IPv4-only network (like most of us), this enables communication with IPv6 hosts quickly, easily, and transparently.

(this HOWTO is archived at http://yyz.us/ipv6-fc2-howto.html)

Simple setup:

1) Append to /etc/sysconfig/network

NETWORKING_IPV6=yes
IPV6_DEFAULTDEV=tun6to4

2) Append to /etc/sysconfig/ifcfg-eth0

IPV6INIT=yes
IPV6TO4INIT=yes

3) Reboot or restart your network interface.

That's it!

If you have an iptables ipv4 firewall, you'll want to

F1) allow ipv6 tunnelled packets to pass through to ip6tables, by allowing protocol 41

iptables -A block -p 41 -j ACCEPT
("block" is a custom chain on my firewall)

F2) duplicate your ipv4 firewall rules for ipv6, using ip6tables. Some things, like masquerade, are not applicable to ipv6.

If you have an ipv4 NATing firewall, which serves as a router for a local network, you'll want to set up radvd and routing rules, so that your hosts autoconfigure ipv6 automatically based on your router's advertisements, and also so that your hosts truly speak native ipv6 without tunneling [the router does the tunnel wrap/unwrap].

R1.1) in /etc/radvd.conf, set "interface ethX" to reflect your router's local LAN interface (eth1 on my own firewall).

R1.2) in radvd.conf, comment out "example of a standard prefix" prefix {} block

R1.3) in radvd.conf, edit the line "prefix 0:0:0:1234::/64" and change "1234" to a network number of your choice.

R1.4) in radvd.conf, edit line "Base6to4Interface ppp0" to reflect the interface doing the 6to4 tunnelling (eth0 on my own firewall).

R2) add routing rules for the local network.

	# ip -6 route add 2002:184a:9ba9:1010::/64 dev eth1
	# ip -6 addr add 2002:184a:9ba9:1010::1 dev eth1

You cat get the 2002:... address (your 6to4 address, formed from your ipv4 address) from your ifconfig. In this example, "eth1" is my local LAN interface. eth0 is the interface to my ISP (DSL modem).

Here is what my ifconfig output looks like, after everything is set up on my router/firewall:

eth0 Link encap:Ethernet HWaddr 00:00:21:DE:DE:B5

          inet addr:24.74.155.XXX  Bcast:255.255.255.255 
Mask:255.255.248.0
          inet6 addr: fe80::200:21ff:fede:deb5/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:8759136 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2238155 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1647432957 (1571.1 Mb)  TX bytes:166256535 (158.5 Mb)
          Interrupt:209 Base address:0x8c00

eth1 Link encap:Ethernet HWaddr 00:C0:9F:39:CD:B0

          inet addr:10.10.10.1  Bcast:10.10.10.255  Mask:255.255.255.0
          inet6 addr: 2002:184a:9ab9:110::1/128 Scope:Global
          inet6 addr: 2002:184a:9ab9:110:2c0:9fff:fe39:cdb0/64 Scope:Global
          inet6 addr: fe80::2c0:9fff:fe39:cdb0/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:9073144 errors:0 dropped:0 overruns:0 frame:0
          TX packets:10916350 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1820645725 (1736.3 Mb)  TX bytes:3611957866 (3444.6
          Base address:0xece0 Memory:fe3e0000-fe400000

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:440 errors:0 dropped:0 overruns:0 frame:0 TX packets:440 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0
RX bytes:54209 (52.9 Kb) TX bytes:54209 (52.9 Kb)

tun6to4 Link encap:IPv6-in-IPv4
inet6 addr: 2002:184a:9ab9::1/16 Scope:Global UP RUNNING NOARP MTU:1480 Metric:1
RX packets:1520 errors:0 dropped:0 overruns:0 frame:0 TX packets:1614 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0
RX bytes:886384 (865.6 Kb) TX bytes:224041 (218.7 Kb)

#1 lore raw

You can also get an IPv6 tunnel at http://tunnelbroker.ipv6.net.au/ that will give you your own IPv6 tunnel and allocation. This isn't just an Australian site either - you can get either AU or US tunnels, and you can apply and use it anywhere in the world.

Cheers,

Paul Fraser
[email protected]

Quoted 101 lines · Jeff Garzik

So, thanks to David Woodhouse for showing me how to do this. IPv6 appears to be very, very close to a Just Works(tm) state.

These instructions are for Fedora Core 2 users, and describe how to set up IPv6 automatically tunnelling (6to4) on an IPv4 network. If you are stuck on an IPv4-only network (like most of us), this enables communication with IPv6 hosts quickly, easily, and transparently.

(this HOWTO is archived at http://yyz.us/ipv6-fc2-howto.html)

Simple setup:

1) Append to /etc/sysconfig/network

NETWORKING_IPV6=yes
IPV6_DEFAULTDEV=tun6to4

2) Append to /etc/sysconfig/ifcfg-eth0

IPV6INIT=yes
IPV6TO4INIT=yes

3) Reboot or restart your network interface.

That's it!

If you have an iptables ipv4 firewall, you'll want to

F1) allow ipv6 tunnelled packets to pass through to ip6tables, by allowing protocol 41

iptables -A block -p 41 -j ACCEPT
("block" is a custom chain on my firewall)

F2) duplicate your ipv4 firewall rules for ipv6, using ip6tables. Some things, like masquerade, are not applicable to ipv6.

If you have an ipv4 NATing firewall, which serves as a router for a local network, you'll want to set up radvd and routing rules, so that your hosts autoconfigure ipv6 automatically based on your router's advertisements, and also so that your hosts truly speak native ipv6 without tunneling [the router does the tunnel wrap/unwrap].

R1.1) in /etc/radvd.conf, set "interface ethX" to reflect your router's local LAN interface (eth1 on my own firewall).

R1.2) in radvd.conf, comment out "example of a standard prefix" prefix {} block

R1.3) in radvd.conf, edit the line "prefix 0:0:0:1234::/64" and change "1234" to a network number of your choice.

R1.4) in radvd.conf, edit line "Base6to4Interface ppp0" to reflect the interface doing the 6to4 tunnelling (eth0 on my own firewall).

R2) add routing rules for the local network.

    # ip -6 route add 2002:184a:9ba9:1010::/64 dev eth1
    # ip -6 addr add 2002:184a:9ba9:1010::1 dev eth1

You cat get the 2002:... address (your 6to4 address, formed from your ipv4 address) from your ifconfig. In this example, "eth1" is my local LAN interface. eth0 is the interface to my ISP (DSL modem).

Here is what my ifconfig output looks like, after everything is set up on my router/firewall:

eth0 Link encap:Ethernet HWaddr 00:00:21:DE:DE:B5

          inet addr:24.74.155.XXX  Bcast:255.255.255.255 Mask:255.255.248.0
          inet6 addr: fe80::200:21ff:fede:deb5/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:8759136 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2238155 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1647432957 (1571.1 Mb)  TX bytes:166256535 (158.5 Mb)
          Interrupt:209 Base address:0x8c00

eth1 Link encap:Ethernet HWaddr 00:C0:9F:39:CD:B0

          inet addr:10.10.10.1  Bcast:10.10.10.255  Mask:255.255.255.0
          inet6 addr: 2002:184a:9ab9:110::1/128 Scope:Global
          inet6 addr: 2002:184a:9ab9:110:2c0:9fff:fe39:cdb0/64 Scope:Global
          inet6 addr: fe80::2c0:9fff:fe39:cdb0/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:9073144 errors:0 dropped:0 overruns:0 frame:0
          TX packets:10916350 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1820645725 (1736.3 Mb)  TX bytes:3611957866 (3444.6
          Base address:0xece0 Memory:fe3e0000-fe400000

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:440 errors:0 dropped:0 overruns:0 frame:0 TX packets:440 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0
RX bytes:54209 (52.9 Kb) TX bytes:54209 (52.9 Kb)

tun6to4 Link encap:IPv6-in-IPv4
inet6 addr: 2002:184a:9ab9::1/16 Scope:Global UP RUNNING NOARP MTU:1480 Metric:1
RX packets:1520 errors:0 dropped:0 overruns:0 frame:0 TX packets:1614 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0
RX bytes:886384 (865.6 Kb) TX bytes:224041 (218.7 Kb)

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#2 lore raw
Quoted 4 lines · Paul Fraser

You can also get an IPv6 tunnel at http://tunnelbroker.ipv6.net.au/ that will give you your own IPv6 tunnel and allocation. This isn't just an Australian site either - you can get either AU or US tunnels, and you can apply and use it anywhere in the world.

Quite a few ISP's will do actual IPv6 or tunnels. In the UK anyone with ADSL can switch provider to aaisp.net and get an IPv6 range and IPv6 tunnel for example.

Alan

#3 lore raw
In reply to Paul Fraser
Quoted 1 line · Paul Fraser

You can also get an IPv6 tunnel at http://tunnelbroker.ipv6.net.au/ that will give you your own IPv6 tunnel and allocation. This isn't just an Australian site either - you can get either AU or US tunnels, and you can apply and use it anywhere in the world.

Agreed, but the point of my post (and 6to4 in general) is that -- like the general theme of IPv6 itself -- using 6to4 under Fedora Core 2 works automatically. No need to worry about logging into a tunnel broker, and no need to worry about traffic with other 6to4 sites being sub-optimally routed.

Jeff

#4 lore raw
In reply to Paul Fraser
Quoted 4 lines · Paul Fraser

You can also get an IPv6 tunnel at http://tunnelbroker.ipv6.net.au/ that will give you your own IPv6 tunnel and allocation. This isn't just an Australian site either - you can get either AU or US tunnels, and you can apply and use it anywhere in the world.

Best way is still asking your ISP for IPv6 address range.

Signature · 2 lines
Tomasz Torcz                Only gods can safely risk perfection,     
[email protected]     it's a dangerous thing for a man.  -- Alia
#5 lore raw
In reply to Jeff Garzik
Quoted 10 lines · In article <[email protected]> you wrote:

If you have an iptables ipv4 firewall, you'll want to

F1) allow ipv6 tunnelled packets to pass through to ip6tables, by allowing protocol 41

iptables -A block -p 41 -j ACCEPT
("block" is a custom chain on my firewall)

F2) duplicate your ipv4 firewall rules for ipv6, using ip6tables. Some things, like masquerade, are not applicable to ipv6.

Note that you have to terminate the tunnel on your firewall in order to filter the encapsulated ipv6. This is important, since letting tunnel packets pass your firewall is a major security problem, otherwise.

Greetings
Bernd

Signature · 2 lines
eckes privat - http://www.eckes.org/
Project Freefire - http://www.freefire.org/
#6 lore raw
Quoted 10 lines · Bernd Eckenfels
Quoted 8 lines · In article <[email protected]> you wrote:

If you have an iptables ipv4 firewall, you'll want to

F1) allow ipv6 tunnelled packets to pass through to ip6tables, by allowing protocol 41

iptables -A block -p 41 -j ACCEPT
("block" is a custom chain on my firewall)

Quoted 7 lines
Quoted 2 lines

F2) duplicate your ipv4 firewall rules for ipv6, using ip6tables. Some things, like masquerade, are not applicable to ipv6.

Note that you have to terminate the tunnel on your firewall in order to filter the encapsulated ipv6. This is important, since letting tunnel packets pass your firewall is a major security problem, otherwise.

And what exactly does this mean?

"terminate the tunnel on your firewall" ???

Would you enlighten me (and the list) how do you do that with ip{,6}tables?

Kalin.

Signature · 3 lines
 || ~~~~~~~~~~~~~~~~~~~~~~ ||
(  ) http://ThinRope.net/ (  )
 || ______________________ ||
#7 lore raw
Quoted 3 lines · In article <[email protected]> you wrote:

And what exactly does this mean?

"terminate the tunnel on your firewall" ???

it means that you should not forward the tunnel packets to internal hosts but to configure the tunnel endpoint interface on (or before) the firewall.

Quoted 1 line

Would you enlighten me (and the list) how do you do that with ip{,6}tables?

Then you can run normal rules on the decapsulated packets.

Gruss
Bernd

Signature · 2 lines
eckes privat - http://www.eckes.org/
Project Freefire - http://www.freefire.org/
#8 lore raw
In reply to Jeff Garzik

Jeff Garzik [2004-08-23 00:51 CEST]:

Quoted 10 lines

So, thanks to David Woodhouse for showing me how to do this. IPv6 appears to be very, very close to a Just Works(tm) state.

These instructions are for Fedora Core 2 users, and describe how to set up IPv6 automatically tunnelling (6to4) on an IPv4 network. If you are stuck on an IPv4-only network (like most of us), this enables communication with IPv6 hosts quickly, easily, and transparently.

(this HOWTO is archived at http://yyz.us/ipv6-fc2-howto.html)

[...]

Quoted 3 lines

If you have an iptables ipv4 firewall, you'll want to

[...]

Quoted 3 lines

F2) duplicate your ipv4 firewall rules for ipv6, using ip6tables. Some things, like masquerade, are not applicable to ipv6.

Does Fedora Core 2 has patches, which add ipv6 statefull inspection to netfilter? If not, this is only true for rules without statefull inspection.

Frank.

Signature · 4 lines
Frank Matthieß

"Was kommt denn als nächstes? Ein Studiengang 6 Semester sendmail mit
 Aufbaustudium `sendmail security'?" -- Frank Guthausen