Docs update for link functionality in user-defined networks

Signed-off-by: Madhu Venugopal <madhu@docker.com>

docs/reference/commandline/network_connect.md

@@ -15,6 +15,9 @@ parent = "smn_cli"
     Connects a container to a network
 
       --help             Print usage
+      --ip               IP Address
+      --ip6              IPv6 Address
+      --link=[]          Add a link to another container
 
 Connects a container to a network. You can connect a container by name
 or by ID. Once connected, the container can communicate with other containers in
@@ -33,7 +36,13 @@ $ docker run -itd --net=multi-host-network busybox
 You can specify the IP address you want to be assigned to the container's interface.
 
 ```bash
-$ docker network connect multi-host-network --ip 10.10.36.122 container2
+$ docker network connect --ip 10.10.36.122 multi-host-network container2
+```
+
+You can use `--link` option to link another container with a prefered alias
+
+```bash
+$ docker network connect --link container1:c1 multi-host-network container2
 ```
 
 You can pause, restart, and stop containers that are connected to a network.
@@ -60,3 +69,4 @@ You can connect a container to one or more networks. The networks need not be th
 * [network ls](network_ls.md)
 * [network rm](network_rm.md)
 * [Understand Docker container networks](../../userguide/networking/dockernetworks.md)
+* [Work with networks](../../userguide/networking/work-with-networks.md)

docs/reference/run.md

@@ -1301,12 +1301,12 @@ specifies `EXPOSE 80` in the Dockerfile). At runtime, the port might be
 bound to 42800 on the host. To find the mapping between the host ports
 and the exposed ports, use `docker port`.
 
-If the operator uses `--link` when starting a new client container, then the
-client container can access the exposed port via a private networking interface.
-Linking is a legacy feature that is only supported on the default bridge
-network. You should prefer the Docker networks feature instead. For more
-information on this feature, see the [*Docker network
-overview*""](../userguide/networking/index.md)).
+If the operator uses `--link` when starting a new client container in the
+default bridge network, then the client container can access the exposed
+port via a private networking interface.
+If `--link` is used when starting a container in a user-defined network as
+described in [*Docker network overview*""](../userguide/networking/index.md)),
+it will provide a named alias for the container being linked to.
 
 ### ENV (environment variables)
 

docs/userguide/networking/default_network/dockerlinks.md

@@ -17,14 +17,11 @@ Before the [Docker networks feature](../dockernetworks.md), you could use the
 Docker link feature to allow containers to discover each other and securely
 transfer information about one container to another container. With the
 introduction of the Docker networks feature, you can still create links but they
-are only supported on the default `bridge` network named `bridge` and appearing
-in your network stack as `docker0`.
+behave differently between default `bridge` network and
+[user defined networks](../work-with-networks.md#linking-containers-in-user-defined-networks)
 
 This section briefly discusses connecting via a network port and then goes into
-detail on container linking. While links are still supported on Docker's default
-network (`bridge`), you should avoid them in preference of the Docker
-networks feature. Linking is expected to be deprecated and removed in a future
-release.
+detail on container linking in default `bridge` network.
 
 ## Connect using network port mapping
 

docs/userguide/networking/work-with-networks.md

@@ -268,7 +268,7 @@ PING container3 (172.25.3.3): 56 data bytes
 round-trip min/avg/max = 0.070/0.081/0.097 ms
 ```
 
-This isn't the case for the default bridge network. Both `container2` and  `container1` are connected to the default bridge network. Docker does not support automatic service discovery on this network. For this reason, pinging  `container1` by name fails as you would expect based on the `/etc/hosts` file:
+This isn't the case for the default `bridge` network. Both `container2` and  `container1` are connected to the default bridge network. Docker does not support automatic service discovery on this network. For this reason, pinging  `container1` by name fails as you would expect based on the `/etc/hosts` file:
 
 ```bash
 / # ping -w 4 container1
@@ -314,6 +314,192 @@ PING 172.17.0.2 (172.17.0.2): 56 data bytes
 You can connect both running and non-running containers to a network. However,
 `docker network inspect` only displays information on running containers.
 
+### Linking containers in user-defined networks
+
+In the above example, container_2 was able to resolve container_3's name automatically
+in the user defined network `isolated_nw`, but the name resolution did not succeed
+automatically in the default `bridge` network. This is expected in order to maintain
+backward compatibility with [legacy link](default_network/dockerlinks.md).
+
+The `legacy link` provided 4 major functionalities to the default `bridge` network.
+
+* name resolution
+* name alias for the linked container using `--link=CONTAINER-NAME:ALIAS`
+* secured container connectivity (in isolation via `--icc=false`)
+* environment variable injection
+
+Comparing the above 4 functionalities with the non-default user-defined networks such as
+`isolated_nw` in this example, without any additional config, `docker network` provides
+
+* automatic name resolution using DNS
+* automatic secured isolated environment for the containers in a network
+* ability to dynamically attach and detach to multiple networks
+* supports the `--link` option to provide name alias for the linked container
+
+Continuing with the above example, create another container `container_4` in `isolated_nw`
+with `--link` to provide additional name resolution using alias for other containers in
+the same network.
+
+```bash
+$ docker run --net=isolated_nw -itd --name=container4 --link container5:c5 busybox
+01b5df970834b77a9eadbaff39051f237957bd35c4c56f11193e0594cfd5117c
+```
+
+With the help of `--link` container4 will be able to reach container5 using the
+aliased name `c5` as well.
+
+Please note that while creating container4, we linked to a container named `container5`
+which is not created yet. That is one of the differences in behavior between the
+`legacy link` in default `bridge` network and the new `link` functionality in user defined
+networks. The `legacy link` is static in nature and it hard-binds the container with the
+alias and it doesnt tolerate linked container restarts. While the new `link` functionality
+in user defined networks are dynamic in nature and supports linked container restarts
+including tolerating ip-address changes on the linked container.
+
+Now let us launch another container named `container5` linking container4 to c4.
+
+```bash
+$ docker run --net=isolated_nw -itd --name=container5 --link container4:c4 busybox
+72eccf2208336f31e9e33ba327734125af00d1e1d2657878e2ee8154fbb23c7a
+```
+
+As expected, container4 will be able to reach container5 by both its container name and
+its alias c5 and container5 will be able to reach container4 by its container name and
+its alias c4.
+
+```bash
+$ docker attach container4
+/ # ping -w 4 c5
+PING c5 (172.25.0.5): 56 data bytes
+64 bytes from 172.25.0.5: seq=0 ttl=64 time=0.070 ms
+64 bytes from 172.25.0.5: seq=1 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=2 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=3 ttl=64 time=0.097 ms
+
+--- c5 ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.070/0.081/0.097 ms
+
+/ # ping -w 4 container5
+PING container5 (172.25.0.5): 56 data bytes
+64 bytes from 172.25.0.5: seq=0 ttl=64 time=0.070 ms
+64 bytes from 172.25.0.5: seq=1 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=2 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=3 ttl=64 time=0.097 ms
+
+--- container5 ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.070/0.081/0.097 ms
+```
+
+```bash
+$ docker attach container5
+/ # ping -w 4 c4
+PING c4 (172.25.0.4): 56 data bytes
+64 bytes from 172.25.0.4: seq=0 ttl=64 time=0.065 ms
+64 bytes from 172.25.0.4: seq=1 ttl=64 time=0.070 ms
+64 bytes from 172.25.0.4: seq=2 ttl=64 time=0.067 ms
+64 bytes from 172.25.0.4: seq=3 ttl=64 time=0.082 ms
+
+--- c4 ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.065/0.070/0.082 ms
+
+/ # ping -w 4 container4
+PING container4 (172.25.0.4): 56 data bytes
+64 bytes from 172.25.0.4: seq=0 ttl=64 time=0.065 ms
+64 bytes from 172.25.0.4: seq=1 ttl=64 time=0.070 ms
+64 bytes from 172.25.0.4: seq=2 ttl=64 time=0.067 ms
+64 bytes from 172.25.0.4: seq=3 ttl=64 time=0.082 ms
+
+--- container4 ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.065/0.070/0.082 ms
+```
+
+Similar to the legacy link functionality the new link alias is localized to a container
+and the aliased name has no meaning outside of the container using the `--link`.
+
+Also, it is important to note that if a container belongs to multiple networks, the
+linked alias is scoped within a given network. Hence the containers can be linked to
+different aliases in different networks.
+
+Extending the example, let us create another network named `local_alias`
+
+```bash
+$ docker network create -d bridge --subnet 172.26.0.0/24 local_alias
+76b7dc932e037589e6553f59f76008e5b76fa069638cd39776b890607f567aaa
+```
+
+let us connect container4 and container5 to the new network `local_alias`
+
+```
+$ docker network connect --link container5:foo local_alias container4
+$ docker network connect --link container4:bar local_alias container5
+```
+
+```bash
+$ docker attach container4
+
+/ # ping -w 4 foo
+PING foo (172.26.0.3): 56 data bytes
+64 bytes from 172.26.0.3: seq=0 ttl=64 time=0.070 ms
+64 bytes from 172.26.0.3: seq=1 ttl=64 time=0.080 ms
+64 bytes from 172.26.0.3: seq=2 ttl=64 time=0.080 ms
+64 bytes from 172.26.0.3: seq=3 ttl=64 time=0.097 ms
+
+--- foo ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.070/0.081/0.097 ms
+
+/ # ping -w 4 c5
+PING c5 (172.25.0.5): 56 data bytes
+64 bytes from 172.25.0.5: seq=0 ttl=64 time=0.070 ms
+64 bytes from 172.25.0.5: seq=1 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=2 ttl=64 time=0.080 ms
+64 bytes from 172.25.0.5: seq=3 ttl=64 time=0.097 ms
+
+--- c5 ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.070/0.081/0.097 ms
+```
+
+Note that the ping succeeds for both the aliases but on different networks.
+Let us conclude this section by disconnecting container5 from the `isolated_nw`
+and observe the results
+
+```
+$ docker network disconnect isolated_nw container5
+
+$ docker attach container4
+
+/ # ping -w 4 c5
+ping: bad address 'c5'
+
+/ # ping -w 4 foo
+PING foo (172.26.0.3): 56 data bytes
+64 bytes from 172.26.0.3: seq=0 ttl=64 time=0.070 ms
+64 bytes from 172.26.0.3: seq=1 ttl=64 time=0.080 ms
+64 bytes from 172.26.0.3: seq=2 ttl=64 time=0.080 ms
+64 bytes from 172.26.0.3: seq=3 ttl=64 time=0.097 ms
+
+--- foo ping statistics ---
+4 packets transmitted, 4 packets received, 0% packet loss
+round-trip min/avg/max = 0.070/0.081/0.097 ms
+
+```
+
+In conclusion, the new link functionality in user defined networks provides all the
+benefits of legacy links while avoiding most of the well-known issues with `legacy links`.
+
+One notable missing functionality compared to `legacy links` is the injection of
+environment variables. Though very useful, environment variable injection is static
+in nature and must be injected when the container is started. One cannot inject
+environment variables into a running container without significant effort and hence
+it is not compatible with `docker network` which provides a dynamic way to connect/
+disconnect containers to/from a network.
+
+
 ## Disconnecting containers
 
 You can disconnect a container from a network using the `docker network