summaryrefslogtreecommitdiffstats
path: root/docs/setupservicevm
diff options
context:
space:
mode:
Diffstat (limited to 'docs/setupservicevm')
-rw-r--r--docs/setupservicevm/2-ipv6-configguide-os-controller.rst5
-rw-r--r--docs/setupservicevm/3-ipv6-configguide-os-compute.rst8
-rw-r--r--docs/setupservicevm/4-ipv6-configguide-servicevm.rst324
-rw-r--r--docs/setupservicevm/images/ipv6-sample-in-horizon.pngbin0 -> 37529 bytes
-rw-r--r--docs/setupservicevm/index.rst1
5 files changed, 330 insertions, 8 deletions
diff --git a/docs/setupservicevm/2-ipv6-configguide-os-controller.rst b/docs/setupservicevm/2-ipv6-configguide-os-controller.rst
index be767f9..f6224e8 100644
--- a/docs/setupservicevm/2-ipv6-configguide-os-controller.rst
+++ b/docs/setupservicevm/2-ipv6-configguide-os-controller.rst
@@ -29,10 +29,9 @@ For exemplary purpose, we assume:
``http://fpaste.org/276949/39476214/``
- **Note 1**: you need to change the IP address of ``ODL_MGR_IP`` to point to your actual IP address
+* Note 1: you need to change the IP address of ``ODL_MGR_IP`` to point to your actual IP address
of Open Daylight Controller.
-
- **Note 2**: You may have to change the value of ``ODL_PROVIDER_MAPPINGS`` and ``PUBLIC_INTERFACE``
+* Note 2: You may have to change the value of ``ODL_PROVIDER_MAPPINGS`` and ``PUBLIC_INTERFACE``
to match your actual network interfaces.
**OS-N-6**: Initiate Openstack setup by invoking ``stack.sh``
diff --git a/docs/setupservicevm/3-ipv6-configguide-os-compute.rst b/docs/setupservicevm/3-ipv6-configguide-os-compute.rst
index c7042a3..d4208bc 100644
--- a/docs/setupservicevm/3-ipv6-configguide-os-compute.rst
+++ b/docs/setupservicevm/3-ipv6-configguide-os-compute.rst
@@ -29,13 +29,11 @@ For exemplary purpose, we assume:
``http://fpaste.org/276958/44395955/``
- **Note 1**: you need to change the IP address of ``SERVICE_HOST`` to point to your actual IP address
+*Note 1: you need to change the IP address of ``SERVICE_HOST`` to point to your actual IP address
of OpenStack Controller.
-
- **Note 2**: you need to change the IP address of ``ODL_MGR_IP`` to point to your actual IP address
+*Note 2: you need to change the IP address of ``ODL_MGR_IP`` to point to your actual IP address
of Open Daylight Controller.
-
- **Note 3**: You may have to change the value of ``ODL_PROVIDER_MAPPINGS`` and ``PUBLIC_INTERFACE``
+*Note 3: You may have to change the value of ``ODL_PROVIDER_MAPPINGS`` and ``PUBLIC_INTERFACE``
to match your actual network interface.
**OS-M-6**: Initiate Openstack setup by invoking ``stack.sh``
diff --git a/docs/setupservicevm/4-ipv6-configguide-servicevm.rst b/docs/setupservicevm/4-ipv6-configguide-servicevm.rst
new file mode 100644
index 0000000..1ffbc53
--- /dev/null
+++ b/docs/setupservicevm/4-ipv6-configguide-servicevm.rst
@@ -0,0 +1,324 @@
+==========================================
+Setting Up a Service VM as an IPv6 vRouter
+==========================================
+
+Now we can start to set up a service VM as an IPv6 vRouter. For exemplary purpose, we assume:
+
+* The hostname of Open Daylight Controller Node is ``opnfv-odl-controller``
+* The hostname of OpenStack Controller Node is ``opnfv-os-controller``
+* The hostname of OpenStack Compute Node is ``opnfv-os-compute``
+* We use ``opnfv`` as username to login.
+* We use ``devstack`` to install OpenStack Kilo, and the directory is ``~/devstack``
+* Note: all IP addresses as shown below are for exemplary purpose.
+
+***************************************************
+Source the Credentials in OpenStack Controller Node
+***************************************************
+
+**SETUP-SVM-1**: Login with username ``opnfv`` in OpenStack Controller Node ``opnfv-os-controller``.
+Start a new terminal, and change directory to where OpenStack is installed.
+
+ ``cd ~/devstack``
+
+**SETUP-SVM-2**: Source the credentials.
+
+ ``opnfv@opnfv-os-controller:~/devstack$ source openrc admin demo``
+
+**************************************
+Add External Connectivity to ``br-ex``
+**************************************
+
+Because we need to manually create networks/subnets to achieve the IPv6 vRouter, we have used the flag
+``NEUTRON_CREATE_INITIAL_NETWORKS=False`` in ``local.conf`` file. When this flag is set to False,
+``devstack`` does not create any networks/subnets during the setup phase.
+
+In OpenStack Controller Node ``opnfv-os-controller``, ``eth1`` is configured to provide external/public connectivity
+for both IPv4 and IPv6. So let us add this interface to ``br-ex`` and move the IP address, including the default route
+from ``eth1`` to ``br-ex``.
+
+**SETUP-SVM-3**: Add ``eth1`` to ``br-ex`` and move the IP address and the default route from ``eth1`` to ``br-ex``
+
+ ``sudo ip addr del <External IP address of opnfv-os-controller> dev eth1 && sudo ovs-vsctl add-port br-ex eth1 &&
+sudo ifconfig eth1 up && sudo ip addr add <External IP address of opnfv-os-controller> dev br-ex && sudo ifconfig
+br-ex up && sudo ip route add default via <Default gateway IP address of opnfv-os-controller> dev br-ex``
+
+* Note: This can be automated in /etc/network/interfaces.
+
+**SETUP-SVM-4**: Verify that ``br-ex`` now has the original external IP address, and that the default route is on
+``br-ex``
+
+ ``opnfv@opnfv-os-controller:~/devstack$ ip a s br-ex``
+| 38: br-ex: <BROADCAST,UP,LOWER_UP> mtu 1430 qdisc noqueue state UNKNOWN group default
+| link/ether 00:50:56:82:42:d1 brd ff:ff:ff:ff:ff:ff
+| inet 198.59.156.113/24 brd 198.59.156.255 scope global br-ex
+| valid_lft forever preferred_lft forever
+| inet6 fe80::543e:28ff:fe70:4426/64 scope link
+| valid_lft forever preferred_lft forever
+
+ ``opnfv@opnfv-os-controller:~/devstack$ ip route``
+| default via 198.59.156.1 dev br-ex
+| 10.134.156.0/24 dev eth0 proto kernel scope link src 10.134.156.113
+| 192.168.122.0/24 dev virbr0 proto kernel scope link src 192.168.122.1
+| 198.59.156.0/24 dev br-ex proto kernel scope link src 198.59.156.113
+
+* Note: The IP addresses above are exemplary purpose
+
+********************************************************
+Create IPv4 Subnet and Router with External Connectivity
+********************************************************
+
+**SETUP-SVM-5**: Create a Neutron router ``ipv4-router`` which needs to provide external connectivity.
+
+ ``neutron router-create ipv4-router``
+
+**SETUP-SVM-6**: Create an external network/subnet ``ext-net`` using the appropriate values based on the
+data-center physical network setup.
+
+ ``neutron net-create --router:external ext-net``
+
+ ``neutron subnet-create --disable-dhcp --allocation-pool start=198.59.156.251,end=198.59.156.254 --gateway
+198.59.156.1 ext-net 198.59.156.0/24``
+
+* Note: The IP addresses in the command above are for exemplary purpose. **Please replace the IP addresses of
+your actual network**.
+
+**SETUP-SVM-7**: Associate the ``ext-net`` to the Neutron router ``ipv4-router``.
+
+ ``neutron router-gateway-set ipv4-router ext-net``
+
+**SETUP-SVM-8**: Create an internal/tenant IPv4 network ``ipv4-int-network1``
+
+ ``neutron net-create ipv4-int-network1``
+
+**SETUP-SVM-9**: Create an IPv4 subnet ``ipv4-int-subnet1`` in the internal network ``ipv4-int-network1``
+
+ ``neutron subnet-create --name ipv4-int-subnet1 --dns-nameserver 8.8.8.8 ipv4-int-network1 20.0.0.0/24``
+
+* Note: The IP addresses in the command above are for exemplary purpose. **Please replace the IP addresses of your
+actual network**
+
+**SETUP-SVM-10**: Associate the IPv4 internal subnet ``ipv4-int-subnet1`` to the Neutron router ``ipv4-router``.
+
+ ``neutron router-interface-add ipv4-router ipv4-int-subnet1``
+
+********************************************************
+Create IPv6 Subnet and Router with External Connectivity
+********************************************************
+
+Now, let us create a second neutron router where we can "manually" spawn a ``radvd`` daemon to simulate an external
+IPv6 router.
+
+**SETUP-SVM-11**: Create a second Neutron router ``ipv6-router`` which needs to provide external connectivity
+
+ ``neutron router-create ipv6-router``
+
+**SETUP-SVM-12**: Associate the ``ext-net`` to the Neutron router ``ipv6-router``
+
+ ``neutron router-gateway-set ipv6-router ext-net``
+
+**SETUP-SVM-13**: Create a second internal/tenant IPv4 network ``ipv4-int-network2``
+
+ ``neutron net-create ipv4-int-network2``
+
+**SETUP-SVM-14**: Create an IPv4 subnet ``ipv4-int-subnet2`` for the ``ipv6-router`` internal network
+``ipv4-int-network2``
+
+ ``neutron subnet-create --name ipv4-int-subnet2 --dns-nameserver 8.8.8.8 ipv4-int-network2 10.0.0.0/24``
+
+* Note: The IP addresses in the command above are for exemplary purpose. **Please replace the IP addresses of
+your actual network**
+
+**SETUP-SVM-15**: Associate the IPv4 internal subnet ``ipv4-int-subnet2`` to the Neutron router ``ipv6-router``.
+
+ ``neutron router-interface-add ipv6-router ipv4-int-subnet2``
+
+**************************************************
+Prepare Image, Metadata and Keypair for Service VM
+**************************************************
+
+**SETUP-SVM-16**: Download ``fedora20`` image which would be used as ``vRouter``
+
+ ``glance image-create --name 'Fedora20' --disk-format qcow2 --container-format bare --is-public true --copy-from http://cloud.fedoraproject.org/fedora-20.x86_64.qcow2``
+
+**SETUP-SVM-17**: Create a keypair
+
+ ``nova keypair-add vRouterKey > ~/vRouterKey``
+
+**SETUP-SVM-18**: Copy the contents from the following url to ``metadata.txt``, i.e. preparing metadata which enables
+IPv6 router functionality inside ``vRouter``
+
+ ``http://fpaste.org/303942/50781923/``
+
+* Note: this ``metadata.txt`` will enable the ``vRouter`` to automatically spawn a ``radvd`` daemon, which advertises
+its IPv6 subnet prefix ``2001:db8:0:2::/64`` in RA (Router Advertisement) message through its ``eth1`` interface to
+other VMs on ``ipv4-int-network1``. The ``radvd`` daemon also advertises the routing information, which routes to
+``2001:db8:0:2::/64`` subnet, in RA (Router Advertisement) message through its ``eth0`` interface to ``eth1``
+interface of ``ipv6-router`` on ``ipv4-int-network2``.
+
+**********************************************************************************************************
+Boot Service VM (``vRouter``) with ``eth0`` on ``ipv4-int-network2`` and ``eth1`` on ``ipv4-int-network1``
+**********************************************************************************************************
+
+Let us boot the service VM (``vRouter``) ``eth0`` interface on ``ipv4-int-network2`` connecting to ``ipv6-router``,
+and ``eth1`` interface on ``ipv4-int-network1`` connecting to ``ipv4-router``.
+
+**SETUP-SVM-19**: Boot the ``vRouter`` using ``Fedora20`` image on the OpenStack Compute Node with hostname
+``opnfv-os-compute``
+
+ ``nova boot --image Fedora20 --flavor m1.small --user-data ./metadata.txt --availability-zone nova:opnfv-os-compute
+--nic net-id=$(neutron net-list | grep -w ipv4-int-network2 | awk '{print $2}')
+--nic net-id=$(neutron net-list | grep -w ipv4-int-network1 | awk '{print $2}') --key-name vRouterKey vRouter``
+
+**SETUP-SVM-20**: Verify that ``Fedora20`` image boots up successfully and the ssh keys are properly injected
+
+ ``nova list``
+
+ ``nova console-log vRouter``
+
+* Note: It may take few minutes for the necessary packages to get installed and ssh keys to be injected.
+
+ ``# Sample Output``
+| [ 762.884523] cloud-init[871]: ec2: #############################################################
+| [ 762.909634] cloud-init[871]: ec2: -----BEGIN SSH HOST KEY FINGERPRINTS-----
+| [ 762.931626] cloud-init[871]: ec2: 2048 e3:dc:3d:4a:bc:b6:b0:77:75:a1:70:a3:d0:2a:47:a9 (RSA)
+| [ 762.957380] cloud-init[871]: ec2: -----END SSH HOST KEY FINGERPRINTS-----
+| [ 762.979554] cloud-init[871]: ec2: #############################################################
+
+*******************************************
+Boot Two Other VMs in ``ipv4-int-network1``
+*******************************************
+
+In order to verify that the setup is working, let us create two cirros VMs with ``eth1`` interface on the
+``ipv4-int-network1``, i.e., connecting to ``vRouter`` ``eth1`` interface for internal network.
+
+We will have to configure appropriate ``mtu`` on the VMs' interface by taking into account the tunneling
+overhead and any physical switch requirements. If so, push the ``mtu`` to the VM either using ``dhcp``
+options or via ``meta-data``.
+
+**SETUP-SVM-21**: Create VM1 on OpenStack Controller Node with hostname ``opnfv-os-controller``
+
+ ``nova boot --image cirros-0.3.4-x86_64-uec --flavor m1.tiny --nic net-id=$(neutron net-list |
+grep -w ipv4-int-network1 | awk '{print $2}')
+--availability-zone nova:opnfv-os-controller --key-name vRouterKey VM1``
+
+**SETUP-SVM-22**: Create VM2 on OpenStack Compute Node with hostname ``opnfv-os-compute``
+
+ ``nova boot --image cirros-0.3.4-x86_64-uec --flavor m1.tiny --nic net-id=$(neutron net-list |
+grep -w ipv4-int-network1 | awk '{print $2}')
+--availability-zone nova:opnfv-os-compute --key-name vRouterKey VM2``
+
+**SETUP-SVM-23**: Confirm that both the VMs are successfully booted.
+
+ ``nova list``
+
+ ``nova console-log VM1``
+
+ ``nova console-log VM2``
+
+**********************************
+Spawn ``RADVD`` in ``ipv6-router``
+**********************************
+
+Let us manually spawn a ``radvd`` daemon inside ``ipv6-router`` namespace to simulate an external router.
+First of all, we will have to identify the ``ipv6-router`` namespace and move to the namespace.
+
+**SETUP-SVM-24**: identify the ``ipv6-router`` namespace and move to the namespace
+
+ ``sudo ip netns exec qrouter-$(neutron router-list | grep -w ipv6-router | awk '{print $2}') bash``
+
+**SETUP-SVM-25**: Upon successful execution of the above command, you will be in the router namespace.
+Now let us configure the IPv6 address on the <qr-xxx> interface.
+
+ ``router_interface=$(ip a s | grep -w "global qr-*" | awk '{print $7}')``
+
+ ``ip -6 addr add 2001:db8:0:1::1 dev $router_interface``
+
+**SETUP-SVM-26**: Copy the following contents to some file, e.g. ``/tmp/br-ex.radvd.conf``
+
+.. code-block::
+
+ interface $router_interface
+ {
+ AdvSendAdvert on;
+ MinRtrAdvInterval 3;
+ MaxRtrAdvInterval 10;
+ prefix 2001:db8:0:1::/64
+ {
+ AdvOnLink on;
+ AdvAutonomous on;
+ };
+ };
+
+**SETUP-SVM-27**: Spawn a ``radvd`` daemon to simulate an external router. This ``radvd`` daemon advertises its
+IPv6 subnet prefix ``2001:db8:0:1::/64`` in RA (Router Advertisement) message through its ``eth1`` interface to
+``eth0`` interface of ``vRouter`` on ``ipv4-int-network2``.
+
+ ``$radvd -C /tmp/br-ex.radvd.conf -p /tmp/br-ex.pid.radvd -m syslog``
+
+**SETUP-SVM-28**: Configure the ``$router_interface`` process entries to process the RA (Router Advertisement)
+message from ``vRouter``, and automatically add a downstream route pointing to the LLA (Link Local Address) of
+``eth0`` interface of the ``vRouter``.
+
+.. code-block::
+
+ sysctl -w net.ipv6.conf.$router_interface.accept_ra=2
+ sysctl -w net.ipv6.conf.$router_interface.accept_ra_rt_info_max_plen=64
+
+**SETUP-SVM-29**: Please note that after the vRouter successfully initializes and starts sending RA (Router
+Advertisement) message (**SETUP-SVM-20**), you would see an IPv6 route to the ''2001:db8:0:2::/64'' prefix
+(subnet) reachable via LLA (Link Local Address) of ``eth0`` interface of the ``vRouter``. You can execute the
+following command to list the IPv6 routes.
+
+ ``ip -6 route show``
+
+********************************
+Testing to Verify Setup Complete
+********************************
+
+Now, let us ``ssh`` to one of the VMs, e.g. VM1, to confirm that it has successfully configured the IPv6 address
+using ``SLAAC`` with prefix ``2001:db8:0:2::/64`` from ``vRouter``.
+
+ * Note: You need to get the IPv4 address associated to VM1. This can be inferred from ``nova list`` command.
+
+**SETUP-SVM-30**: ``ssh`` VM1
+
+ ``ssh -i /home/odl/vRouterKey cirros@<VM1-IPv4-address>``
+
+If everything goes well, ``ssh`` will be successful and you will be logged into VM1. Run some commands to verify
+that IPv6 addresses are configured on ``eth0`` interface.
+
+**SETUP-SVM-31**: Show an IPv6 address with a prefix of ``2001:db8:0:2::/64``
+
+ ``ip address show``
+
+**SETUP-SVM-32**: ping some external IPv6 address, e.g. ``ipv6-router``
+
+ ``ping6 2001:db8:0:1::1``
+
+If the above ping6 command succeeds, it implies that ``vRouter`` was able to successfully forward the IPv6 traffic
+to reach external ``ipv6-router``.
+
+**SETUP-SVM-33**: When all tests show that the setup works as expected, You can now exit the ``ipv6-router`` namespace.
+
+ ``exit``
+
+**********
+Next Steps
+**********
+
+Congratulations, you have completed the setup of using a service VM to act as an IPv6 vRouter. This setup allows further
+open innovation by any 3rd-party. Please refer to relevant sections in User's Guide for further value-added services on
+this IPv6 vRouter.
+
+********************************************************
+Sample Network Topology of this Setup through Horizon UI
+********************************************************
+
+The sample network topology of above setup is shown in Horizon UI as follows :numref:`figure3`:
+
+.. figure:: images/ipv6-sample-in-horizon.png
+ :name: figure3
+ :width: 100%
+
+ Sample Network Topology in Horizon UI
+
diff --git a/docs/setupservicevm/images/ipv6-sample-in-horizon.png b/docs/setupservicevm/images/ipv6-sample-in-horizon.png
new file mode 100644
index 0000000..2eb03f9
--- /dev/null
+++ b/docs/setupservicevm/images/ipv6-sample-in-horizon.png
Binary files differ
diff --git a/docs/setupservicevm/index.rst b/docs/setupservicevm/index.rst
index ea6392e..facc53c 100644
--- a/docs/setupservicevm/index.rst
+++ b/docs/setupservicevm/index.rst
@@ -20,4 +20,5 @@ Setting Up a Service VM as an IPv6 vRouter
1-ipv6-configguide-odl-setup.rst
2-ipv6-configguide-os-controller.rst
3-ipv6-configguide-os-compute.rst
+ 4-ipv6-configguide-servicevm.rst