beenull/moby
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Update IPv6 docs with modern ip command use + sysctl

Browse source 
Fixes: #10701

Updates the IPv6 documentation and images to reflect the more modern
`ip -6` command set versus `route` and `ifconfig`.  Also removes the
use of the special 2002: address range as that is reserved for 6to4
addressing, as well as use of any public address range and re-works the
switched routing example to use 3 subnets of the documentation IPv6
prefix range.

Also conformed all use of addresses to the same doc range per @MalteJ's
commit.

Docker-DCO-1.1-Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com> (github: estesp)
Signed-off-by: Malte Janduda <mail@janduda.net>
This commit is contained in:
Phil Estes 2015-02-12 17:00:30 -05:00
parent df7a04f9a6
commit 95668ed128
9 changed files with 51 additions and 60 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
docs/sources/article-img/ipv6_basic_host_config.gliffy
View file

File diff suppressed because one or more lines are too long

2
docs/sources/article-img/ipv6_basic_host_config.svg
View file

File diff suppressed because one or more lines are too long
Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 19 KiB

After

Width:  |  Height:  |  Size: 30 KiB

2
docs/sources/article-img/ipv6_routed_network_example.gliffy
View file

File diff suppressed because one or more lines are too long

2
docs/sources/article-img/ipv6_routed_network_example.svg
View file

File diff suppressed because one or more lines are too long
Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 80 KiB

After

Width:  |  Height:  |  Size: 173 KiB

2
docs/sources/article-img/ipv6_slash64_subnet_config.gliffy
View file

File diff suppressed because one or more lines are too long

2
docs/sources/article-img/ipv6_slash64_subnet_config.svg
View file

File diff suppressed because one or more lines are too long
Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 30 KiB

After

Width:  |  Height:  |  Size: 74 KiB

2
docs/sources/article-img/ipv6_switched_network_example.gliffy
View file

File diff suppressed because one or more lines are too long

2
docs/sources/article-img/ipv6_switched_network_example.svg
View file

File diff suppressed because one or more lines are too long
Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 72 KiB

After

Width:  |  Height:  |  Size: 175 KiB

95
docs/sources/articles/networking.md
View file

@ -433,7 +433,7 @@ To assign globally routable IPv6 addresses to your containers you have to
specify an IPv6 subnet to pick the addresses from. Set the IPv6 subnet via the specify an IPv6 subnet to pick the addresses from. Set the IPv6 subnet via the
`--fixed-cidr-v6` parameter when starting Docker daemon: `--fixed-cidr-v6` parameter when starting Docker daemon:
docker -d --ipv6 --fixed-cidr-v6="2001:db8:0:2::/64" docker -d --ipv6 --fixed-cidr-v6="2001:db8:1::/64"
The subnet for Docker containers should at least have a size of `/80`. This way The subnet for Docker containers should at least have a size of `/80`. This way
an IPv6 address can end with the container's MAC address and you prevent NDP an IPv6 address can end with the container's MAC address and you prevent NDP
@ -443,11 +443,11 @@ With the `--fixed-cidr-v6` parameter set Docker will add a new route to the
routing table. Further IPv6 routing will be enabled (you may prevent this by routing table. Further IPv6 routing will be enabled (you may prevent this by
starting Docker daemon with `--ip-forward=false`): starting Docker daemon with `--ip-forward=false`):
$ route -A inet6 add 2001:db8:0:2::/64 dev docker0 $ ip -6 route add 2001:db8:1::/64 dev docker0
$ echo 1 > /proc/sys/net/ipv6/conf/default/forwarding $ sysctl net.ipv6.conf.default.forwarding=1
$ echo 1 > /proc/sys/net/ipv6/conf/all/forwarding $ sysctl net.ipv6.conf.all.forwarding=1
All traffic to the subnet `2001:db8:0:2::/64` will now be routed All traffic to the subnet `2001:db8:1::/64` will now be routed
via the `docker0` interface. via the `docker0` interface.
Be aware that IPv6 forwarding may interfere with your existing IPv6 Be aware that IPv6 forwarding may interfere with your existing IPv6
@ -457,7 +457,7 @@ enabled forwarding will result in rejecting Router Advertisements. E.g., if you
want to configure `eth0` via Router Advertisements you should set: want to configure `eth0` via Router Advertisements you should set:
``` ```
$ echo 2 > /proc/sys/net/ipv6/conf/eth0/accept_ra $ sysctl net.ipv6.conf.eth0.accept_ra=2
``` ```
![](/article-img/ipv6_basic_host_config.svg) ![](/article-img/ipv6_basic_host_config.svg)
@ -465,45 +465,36 @@ want to configure `eth0` via Router Advertisements you should set:
Every new container will get an IPv6 address from the defined subnet. Further Every new container will get an IPv6 address from the defined subnet. Further
a default route will be added via the gateway `fe80::1` on `eth0`: a default route will be added via the gateway `fe80::1` on `eth0`:
docker run -it ubuntu bash -c "ifconfig eth0; route -A inet6" docker run -it ubuntu bash -c "ip -6 addr show dev eth0; ip -6 route show"
eth0 Link encap:Ethernet HWaddr 02:42:ac:11:00:02 15: eth0: <BROADCAST,UP,LOWER_UP> mtu 1500
inet addr:172.17.0.2 Bcast:0.0.0.0 Mask:255.255.0.0 inet6 2001:db8:1:0:0:242:ac11:3/64 scope global
inet6 addr: 2001:db8:0:2::1/64 Scope:Global valid_lft forever preferred_lft forever
inet6 addr: fe80::42:acff:fe11:2/64 Scope:Link inet6 fe80::42:acff:fe11:3/64 scope link
UP BROADCAST MTU:1500 Metric:1 valid_lft forever preferred_lft forever
RX packets:1 errors:0 dropped:0 overruns:0 frame:0
TX packets:1 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:110 (110.0 B) TX bytes:110 (110.0 B)
Kernel IPv6 routing table 2001:db8:1::/64 dev eth0 proto kernel metric 256
Destination Next Hop Flag Met Ref Use If fe80::/64 dev eth0 proto kernel metric 256
2001:db8:0:2::/64 :: U 256 0 0 eth0 default via fe80::1 dev eth0 metric 1024
fe80::/64 :: U 256 0 0 eth0
::/0 fe80::1 UG 1024 0 0 eth0
::/0 :: !n -1 1 1 lo
::1/128 :: Un 0 1 0 lo
ff00::/8 :: U 256 1 0 eth0
::/0 :: !n -1 1 1 lo
In this example the Docker container is assigned a link-local address with the In this example the Docker container is assigned a link-local address with the
network suffix `/64` (here: `fe80::42:acff:fe11:2/64`) and a globally routable network suffix `/64` (here: `fe80::42:acff:fe11:3/64`) and a globally routable
IPv6 address (here: `2001:db8:0:2::1/64`). The container will create connections IPv6 address (here: `2001:db8:1:0:0:242:ac11:3/64`). The container will create
to addresses outside of the `2001:db8:0:2::/64` network via the link-local connections to addresses outside of the `2001:db8:1::/64` network via the
gateway at `fe80::1` on `eth0`. link-local gateway at `fe80::1` on `eth0`.
Often servers or virtual machines get a `/64` IPv6 subnet assigned. In this case Often servers or virtual machines get a `/64` IPv6 subnet assigned (e.g.
you can split it up further and provide Docker a `/80` subnet while using a `2001:db8:23:42::/64`). In this case you can split it up further and provide
separate `/80` subnet for other applications on the host: Docker a `/80` subnet while using a separate `/80` subnet for other
applications on the host:
![](/article-img/ipv6_slash64_subnet_config.svg) ![](/article-img/ipv6_slash64_subnet_config.svg)
In this setup the subnet `2001:db8::/80` with a range from `2001:db8::0:0:0:0` In this setup the subnet `2001:db8:23:42::/80` with a range from `2001:db8:23:42:0:0:0:0`
to `2001:db8::0:ffff:ffff:ffff` is attached to `eth0`, with the host listening to `2001:db8:23:42:0:ffff:ffff:ffff` is attached to `eth0`, with the host listening
at `2001:db8::1`. The subnet `2001:db8:0:0:0:1::/80` with an address range from at `2001:db8:23:42::1`. The subnet `2001:db8:23:42:1::/80` with an address range from
`2001:db8::1:0:0:0` to `2001:db8::1:ffff:ffff:ffff` is attached to `docker0` and `2001:db8:23:42:1:0:0:0` to `2001:db8:23:42:1:ffff:ffff:ffff` is attached to
will be used by containers. `docker0` and will be used by containers.
### Docker IPv6 Cluster ### Docker IPv6 Cluster
@ -514,29 +505,29 @@ example:
![](/article-img/ipv6_switched_network_example.svg) ![](/article-img/ipv6_switched_network_example.svg)
The Docker hosts are in the `2000::/64` subnet. Host1 is configured The Docker hosts are in the `2001:db8:0::/64` subnet. Host1 is configured
to provide addresses from the `2001::/64` subnet to its containers. It has three to provide addresses from the `2001:db8:1::/64` subnet to its containers. It
routes configured: has three routes configured:
- Route all traffic to `2000::/64` via `eth0` - Route all traffic to `2001:db8:0::/64` via `eth0`
- Route all traffic to `2001::/64` via `docker0` - Route all traffic to `2001:db8:1::/64` via `docker0`
- Route all traffic to `2002::/64` via Host2 with IP `2000::2` - Route all traffic to `2001:db8:2::/64` via Host2 with IP `2001:db8::2`
Host1 also acts as a router on OSI layer 3. When one of the network clients Host1 also acts as a router on OSI layer 3. When one of the network clients
tries to contact a target that is specified in Host1's routing table Host1 will tries to contact a target that is specified in Host1's routing table Host1 will
forward the traffic accordingly. It acts as a router for all networks it knows: forward the traffic accordingly. It acts as a router for all networks it knows:
`2000:/64`, `2001:/64` and `2002::/64`. `2001:db8::/64`, `2001:db8:1::/64` and `2001:db8:2::/64`.
On Host2 we have nearly the same configuration. Host2's containers will get IPv6 On Host2 we have nearly the same configuration. Host2's containers will get
addresses from `2002::/64`. Host2 has three routes configured: IPv6 addresses from `2001:db8:2::/64`. Host2 has three routes configured:
- Route all traffic to `2000::/64` via `eth0` - Route all traffic to `2001:db8:0::/64` via `eth0`
- Route all traffic to `2002::/64` via `docker0` - Route all traffic to `2001:db8:2::/64` via `docker0`
- Route all traffic to `2001::/64` via Host1 with IP `2000::1` - Route all traffic to `2001:db8:1::/64` via Host1 with IP `2001:db8:0::1`
The difference to Host1 is that the network `2002::/64` is directly attached to The difference to Host1 is that the network `2001:db8:2::/64` is directly
the host via its `docker0` interface whereas it reaches `2001::/64` via Host1's attached to the host via its `docker0` interface whereas it reaches
IPv6 address `2000::1`. `2001:db8:1::/64` via Host1's IPv6 address `2001:db8::1`.
This way every container is able to contact every other container. The This way every container is able to contact every other container. The
containers `Container1-*` share the same subnet and contact each other directly. containers `Container1-*` share the same subnet and contact each other directly.