ligang/vpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: Rework the VPP progressive Tutorial.

Browse Source 
Change-Id: If5b0d07ea90d978c6b1f11210a661876b7929653
Signed-off-by: John DeNisco <jdenisco@cisco.com>
This commit is contained in:
John DeNisco
2018-08-16 13:50:02 -04:00
committed by Damjan Marion
parent a60c3ed14b
commit c4c72d2835
21 changed files with 394 additions and 471 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
docs/gettingstarted/developers/index.rst
View File

@ -1,8 +1,8 @@
.. _gstarteddevel:
##########
Developers
##########
###############
For Developers
###############
The Developers section covers the following areas:

4
docs/gettingstarted/developers/vnet.md
View File

@ -17,10 +17,6 @@ classification, feeds ipv4-input, ipv6-input, arp-input etc.\] and
"ipv4-input-no-checksum" \[if hardware can classify, perform ipv4 header
checksum\].
![image](/_images/VNET_Features.png)
List of features and layer areas that VNET works with:
Effective graph dispatch function coding
----------------------------------------

1
docs/gettingstarted/index.rst
View File

@ -33,6 +33,7 @@ The Writing VPP Documentation section covers the following topics:
.. toctree::
:maxdepth: 2
progressivevpp/index.rst
users/index.rst
developers/index.rst
writingdocs/index.rst

27
docs/gettingstarted/progressivevpp/index.rst Normal file
View File

@ -0,0 +1,27 @@
.. _progressivevpp:
.. toctree::
########################
Progressive VPP Tutorial
########################
Learn to run FD.io VPP on a single Ubuntu 16.04 VM using Vagrant with this walkthrough
covering basic FD.io VPP senarios. Useful FD.io VPP commands will be used, and
will discuss basic operations, and the state of a running FD.io VPP on a system.
.. note::
This is **not** intended to be a 'How to Run in a Production Environment' set of instructions.
For more information on using VPP with Virtual Box/Vagrant, please refer to :ref:`vppvagrant`
.. toctree::
settingupenvironment.rst
runningvpp.rst
interface.rst
traces.rst
twovppinstances.rst
routing.rst
switching.rst

55
docs/guides/progressivevpp/interface.rst → docs/gettingstarted/progressivevpp/interface.rst
View File

@ -2,11 +2,20 @@
.. toctree::
Interface
=========
Creating an Interface
======================
Skills to be Learned
--------------------
#. Create a veth interface in Linux host
#. Assign an IP address to one end of the veth interface in the Linux host
#. Create a vpp host-interface that connected to one end of a veth interface via AF_PACKET
#. Add an ip address to a vpp interface
VPP commands learned in this exercise
--------------------------------------
VPP command learned in this exercise
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#. `create host-interface <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_devices_af_packet.html#clicmd_create_host-interface>`_
#. `set int state <https://docs.fd.io/vpp/17.04/clicmd_src_vnet.html#clicmd_set_interface_state>`_
#. `set int ip address <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_ip.html#clicmd_set_interface_ip_address>`_
@ -20,7 +29,7 @@ VPP command learned in this exercise
#. `show ip fib <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_fib.html#clicmd_show_ip_fib>`_
Topology
~~~~~~~~
---------
.. figure:: /_images/Create_Interface_Topology.jpg
:alt: Figure: Create Interface Topology
@ -28,13 +37,13 @@ Topology
Figure: Create Interface Topology
Initial State
~~~~~~~~~~~~~
--------------
The initial state here is presumed to be the final state from the
exercise `VPP Basics <VPP/Progressive_VPP_Tutorial#Exercise:_vpp_basics>`__
The initial state here is presumed to be the final state from the previous sections
of the tutorial.
Create veth interfaces on host
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-------------------------------
In Linux, there is a type of interface call 'veth'. Think of a 'veth'
interface as being an interface that has two ends to it (rather than
@ -55,7 +64,7 @@ Turn up both ends:
$ sudo ip link set dev vpp1host up
Assign an IP address
~~~~~~~~~~~~~~~~~~~~
---------------------
.. code-block:: console
@ -65,7 +74,7 @@ Display the result:
.. code-block:: console
$ sudo ip addr show vpp1host
$ ip addr show vpp1host
5: vpp1host@vpp1out: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether e2:0f:1e:59:ec:f7 brd ff:ff:ff:ff:ff:ff
inet 10.10.1.1/24 scope global vpp1host
@ -74,7 +83,29 @@ Display the result:
valid_lft forever preferred_lft forever
Create vpp host-interface
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
--------------------------
Make sure VPP is running, if not start it.
.. code-block:: console
$ ps -eaf | grep vpp
vagrant 2141 903 0 05:28 pts/0 00:00:00 grep --color=auto vpp
# vpp is not running, so start it
$ sudo /usr/bin/vpp -c startup1.conf
These commands are run from the vpp shell. Enter the VPP shell with the following
command:
.. code-block:: console
$ sudo vppctl -s /run/vpp/cli-vpp1.sock
_______ _ _ _____ ___
__/ __/ _ \ (_)__ | | / / _ \/ _ \
_/ _// // / / / _ \ | |/ / ___/ ___/
/_/ /____(_)_/\___/ |___/_/ /_/
vpp#
Create a host interface attached to **vpp1out**.

21
docs/guides/progressivevpp/routing.rst → docs/gettingstarted/progressivevpp/routing.rst
View File

@ -6,7 +6,7 @@ Routing
=======
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
---------------------
In this exercise you will learn these new skills:
@ -22,14 +22,14 @@ And revisit the old ones:
#. ping from FD.io VPP to host
#. Examine and clear trace on vpp1 and vpp2
FD.io VPP command learned in this exercise
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
VPP command learned in this exercise
-------------------------------------
#. `ip route
add <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_ip.html#clicmd_ip_route>`__
Topology
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------
.. figure:: /_images/Connecting_two_vpp_instances_with_memif.png
:alt: Connect two FD.io VPP topology
@ -37,14 +37,14 @@ Topology
Connect two FD.io VPP topology
Initial State
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------
The initial state here is presumed to be the final state from the
exercise `Connecting two FD.io VPP
instances <VPP/Progressive_VPP_Tutorial#Connecting_two_vpp_instances>`__
Setup host route
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-----------------
.. code-block:: console
@ -56,14 +56,19 @@ Setup host route
10.10.2.0/24 via 10.10.1.2 dev vpp1host
Setup return route on vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------------------------
.. code-block:: console
$ sudo vppctl -s /run/vpp/cli-vpp2.sock
vpp# ip route add 10.10.1.0/24 via 10.10.2.1
Ping from host through vpp1 to vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
------------------------------------
The connection from vpp1 to vpp2 uses the **memif** driver, the connection to the host
uses the **af-packet** driver. To trace packets from the host we use **af-packet-input** from
vpp1 to vpp2 we use **memif-input**.
#. Setup a trace on vpp1 and vpp2
#. Ping 10.10.2.2 from the host

74
docs/gettingstarted/progressivevpp/runningvpp.rst Normal file
View File

@ -0,0 +1,74 @@
.. _runningvpp:
Running VPP
===========
Using the files we create in :ref`settingupenvironment` we will now start and
run VPP.
VPP runs in userspace. In a production environment you will often run it
with DPDK to connect to real NICs or vhost to connect to VMs. In those
circumstances you usually run a single instance of VPP.
For purposes of this tutorial, it is going to be extremely useful to run
multiple instances of VPP, and connect them to each other to form a
topology. Fortunately, VPP supports this.
Using the files we created in setup we will start VPP.
.. code-block:: console
$ sudo /usr/bin/vpp -c startup1.conf
vlib_plugin_early_init:361: plugin path /usr/lib/vpp_plugins:/usr/lib64/vpp_plugins
load_one_plugin:189: Loaded plugin: abf_plugin.so (ACL based Forwarding)
load_one_plugin:189: Loaded plugin: acl_plugin.so (Access Control Lists)
load_one_plugin:189: Loaded plugin: avf_plugin.so (Intel Adaptive Virtual Function (AVF) Device Plugin)
.........
$
If VPP does not start you can try adding **nodaemon** to the startup.conf file in the
**unix** section. This should provide more information in the output.
startup.conf example with nodaemon:
.. code-block:: console
unix {nodaemon cli-listen /run/vpp/cli-vpp1.sock}
api-segment { prefix vpp1 }
plugins { plugin dpdk_plugin.so { disable } }
The command **vppctl** will launch a VPP shell with which you can run
VPP commands interactively.
We should now be able to execute the VPP shell and show the version.
.. code-block:: console
$ sudo vppctl -s /run/vpp/cli-vpp1.sock
_______ _ _ _____ ___
__/ __/ _ \ (_)__ | | / / _ \/ _ \
_/ _// // / / / _ \ | |/ / ___/ ___/
/_/ /____(_)_/\___/ |___/_/ /_/
vpp# show version
vpp v18.07-release built by root on c469eba2a593 at Mon Jul 30 23:27:03 UTC 2018
vpp#
.. note::
Use ctrl-d or q to exit from the VPP shell.
If you are going to run several instances of VPP this way be sure to kill them
when you are finished.
You can use something like the following:
.. code-block:: console
$ ps -eaf | grep vpp
root 2067 1 2 05:12 ? 00:00:00 /usr/bin/vpp -c startup1.conf
vagrant 2070 903 0 05:12 pts/0 00:00:00 grep --color=auto vpp
$ kill -9 2067
$ ps -eaf | grep vpp
vagrant 2074 903 0 05:13 pts/0 00:00:00 grep --color=auto vpp

154
docs/gettingstarted/progressivevpp/settingupenvironment.rst Normal file
View File

@ -0,0 +1,154 @@
.. _settingupenvironment:
Setting up your environment
===========================
All of these exercises are designed to be performed on an Ubuntu 16.04 (Xenial) box.
* If you have an Ubuntu 16.04 box on which you have sudo or root access, you can feel free to use that.
* If you do not, a Vagrantfile is provided to setup a basic Ubuntu 16.04 box for you in the the steps below.
Install Virtual Box and Vagrant
-------------------------------
You will need to install Virtual Box and Vagrant. If you have not installed Virtual Box or Vagrant please
refer to :ref:`installingVboxVagrant` to install Virtual Box and Vagrant.
Create a Vagrant Directory
---------------------------
To get started create a directory for vagrant
.. code-block:: console
$ mkdir vpp-tutorial
$ cd vpp-tutorial
Create a file called **Vagrantfile** with the following contents:
.. code-block:: ruby
# -*- mode: ruby -*-
# vi: set ft=ruby :
Vagrant.configure(2) do |config|
config.vm.box = "puppetlabs/ubuntu-16.04-64-nocm"
config.vm.box_check_update = false
vmcpu=(ENV['VPP_VAGRANT_VMCPU'] || 2)
vmram=(ENV['VPP_VAGRANT_VMRAM'] || 4096)
config.ssh.forward_agent = true
config.vm.provider "virtualbox" do |vb|
vb.customize ["modifyvm", :id, "--ioapic", "on"]
vb.memory = "#{vmram}"
vb.cpus = "#{vmcpu}"
#support for the SSE4.x instruction is required in some versions of VB.
vb.customize ["setextradata", :id, "VBoxInternal/CPUM/SSE4.1", "1"]
vb.customize ["setextradata", :id, "VBoxInternal/CPUM/SSE4.2", "1"]
end
end
Running Vagrant
---------------
VPP runs in userspace. In a production environment you will often run it with
DPDK to connect to real NICs or vhost to connect to VMs.mIn those circumstances
you usually run a single instance of VPP.
For purposes of this tutorial, it is going to be extremely useful to run multiple
instances of vpp, and connect them to each other to form a topology. Fortunately,
VPP supports this.
When running multiple VPP instances, each instance needs to have specified a 'name'
or 'prefix'. In the example below, the 'name' or 'prefix' is "vpp1". Note that only
one instance can use the dpdk plugin, since this plugin is trying to acquire a lock
on a file.
Setting up VPP environment with Vagrant
---------------------------------------------
After setting up Vagrant, use these commands on your Vagrant directory to boot the VM:
.. code-block:: console
$ vagrant up
$ vagrant ssh
$ sudo apt-get update
$ sudo reboot -n
$ # Wait for the VM to reboot
$ vagrant ssh
Install VPP
------------
Now that the VM is updated, we will install the VPP packages.
For more on installing VPP please refer to :ref:`installingVPP`.
For this tutorial we need to install VPP by modifying the file
**/etc/apt/sources.list.d/99fd.io.list**.
Write this file with the following contents:
.. code-block:: console
deb [trusted=yes] https://nexus.fd.io/content/repositories/fd.io.ubuntu.xenial.main/ ./
Then execute the following commands.
.. code-block:: console
$ sudo bash
# apt-get update
# apt-get install vpp-lib vpp vpp-plugins
#
Stop VPP for this tutorial. We will be creating our own instances of VPP.
.. code-block:: console
# service vpp stop
#
Create some startup files
--------------------------
We will create some startup files for the use of this tutorial. Typically you will
modify the startup.conf file found in /etc/vpp/startup.conf. For more information
on this file refer to :ref:`startup`.
When running multiple VPP instances, each instance needs to have
specified a 'name' or 'prefix'. In the example below, the 'name' or 'prefix'
is "vpp1". Note that only one instance can use the dpdk plugin, since this
plugin is trying to acquire a lock on a file. These startup files we create will
disable the dpdk plugin.
Also in our startup files notice **api-segment**. **api-segment {prefix vpp1}**
tells FD.io VPP how to name the files in /dev/shm/ for your VPP instance
differently from the default. **unix {cli-listen /run/vpp/cli-vpp1.sock}**
tells vpp to use a non-default socket file when being addressed by vppctl.
Now create 2 files named startup1.conf and startup2.conf with the following
content. These files can be located anywhere. We specify the location when we
start VPP.
startup1.conf:
.. code-block:: console
unix {cli-listen /run/vpp/cli-vpp1.sock}
api-segment { prefix vpp1 }
plugins { plugin dpdk_plugin.so { disable } }
startup2.conf:
.. code-block:: console
unix {cli-listen /run/vpp/cli-vpp2.sock}
api-segment { prefix vpp2 }
plugins { plugin dpdk_plugin.so { disable } }

25
docs/guides/progressivevpp/switching.rst → docs/gettingstarted/progressivevpp/switching.rst
View File

@ -6,7 +6,7 @@ Switching
=========
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
----------------------
#. Associate an interface with a bridge domain
#. Create a loopback interaface
@ -14,7 +14,7 @@ Skills to be Learned
#. Examine a bridge domain
FD.io VPP command learned in this exercise
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------------------------------------
#. `show
bridge <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_l2.html#clicmd_show_bridge-domain>`__
@ -26,7 +26,7 @@ FD.io VPP command learned in this exercise
verbose <https://docs.fd.io/vpp/17.04/clicmd_src_vnet_l2.html#clicmd_show_l2fib>`__
Topology
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------
.. figure:: /_images/Switching_Topology.jpg
:alt: Switching Topology
@ -34,7 +34,7 @@ Topology
Switching Topology
Initial state
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------------
Unlike previous exercises, for this one you want to start tabula rasa.
@ -46,16 +46,17 @@ To clear existing config from previous exercises run:
$ ps -ef | grep vpp | awk '{print $2}'| xargs sudo kill
$ sudo ip link del dev vpp1host
$ # do the next command if you are cleaing up from this example
$ sudo ip link del dev vpp1vpp2
Run FD.io VPP instances
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
------------------------
#. Run a vpp instance named **vpp1**
#. Run a vpp instance named **vpp2**
Connect vpp1 to host
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------------------
#. Create a veth with one end named vpp1host and the other named
vpp1out.
@ -63,7 +64,7 @@ Connect vpp1 to host
#. Add ip address 10.10.1.1/24 on vpp1host
Connect vpp1 to vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------------------
#. Create a veth with one end named vpp1vpp2 and the other named
vpp2vpp1.
@ -71,7 +72,7 @@ Connect vpp1 to vpp2
#. Connect vpp2vpp1 to vpp2.
Configure Bridge Domain on vpp1
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------------------------
Check to see what bridge domains already exist, and select the first
bridge domain number not in use:
@ -128,7 +129,7 @@ Examine bridge domain 1:
host-vpp1vpp2 2 1 0 - * none
Configure loopback interface on vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-------------------------------------
.. code-block:: console
@ -139,7 +140,7 @@ Add the ip address 10.10.1.2/24 to vpp2 interface loop0. Set the state
of interface loop0 on vpp2 to 'up'
Configure bridge domain on vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------------------------
Check to see the first available bridge domain ID (it will be 1 in this
case)
@ -160,7 +161,7 @@ Add interface vpp2vpp1 to bridge domain 1
Examine the bridge domain and interfaces.
Ping from host to vpp and vpp to host
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------------------------------
#. Add trace on vpp1 and vpp2
#. ping from host to 10.10.1.2
@ -169,7 +170,7 @@ Ping from host to vpp and vpp to host
#. Examine and clear trace on vpp1 and vpp2
Examine l2 fib
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------------
.. code-block:: console

55
docs/guides/progressivevpp/traces.rst → docs/gettingstarted/progressivevpp/traces.rst
View File

@ -2,11 +2,22 @@
.. toctree::
Traces
======
Using the trace command
========================
Skills to be Learned
---------------------
#. Setup a 'trace'
#. View a 'trace'
#. Clear a 'trace'
#. Verify using ping from host
#. Ping from vpp
#. Examine Arp Table
#. Examine ip fib
Basic Trace Commands
~~~~~~~~~~~~~~~~~~~~
---------------------
Show trace buffer [max COUNT].
@ -14,7 +25,6 @@ Show trace buffer [max COUNT].
vpp# show trace
Clear trace buffer and free memory.
.. code-block:: console
@ -27,26 +37,15 @@ filter trace output - include NODE COUNT | exclude NODE COUNT | none.
vpp# trace filter <include NODE COUNT | exclude NODE COUNT | none>
Skills to be Learned
~~~~~~~~~~~~~~~~~~~~
#. Setup a 'trace'
#. View a 'trace'
#. Clear a 'trace'
#. Verify using ping from host
#. Ping from vpp
#. Examine Arp Table
#. Examine ip fib
Add Trace
~~~~~~~~~
----------
.. code-block:: console
vpp# trace add af-packet-input 10
Ping from Host to FD.io VPP
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ping from Host to VPP
----------------------
.. code-block:: console
@ -59,12 +58,12 @@ Ping from Host to FD.io VPP
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.283/0.283/0.283/0.000 ms
Examine Trace of ping from host to FD.io VPP
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Examine Trace of ping from host to VPP
----------------------------------------
.. code-block:: console
# vppctl
$ sudo vppctl -s /run/vpp/cli-vpp1.sock
vpp# show trace
------------------- Start of thread 0 vpp_main -------------------
Packet 1
@ -121,14 +120,14 @@ Examine Trace of ping from host to FD.io VPP
ICMP echo_reply checksum 0xc843
Clear trace buffer
~~~~~~~~~~~~~~~~~~
-------------------
.. code-block:: console
vpp# clear trace
Ping from FD.io VPP to Host
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ping from VPP to Host
-----------------------
.. code-block:: console
@ -141,8 +140,8 @@ Ping from FD.io VPP to Host
Statistics: 5 sent, 5 received, 0% packet loss
Examine Trace of ping from FD.io VPP to host
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Examine Trace of ping from VPP to host
---------------------------------------
The output will demonstrate FD.io VPP's trace of ping for all packets.
@ -442,7 +441,7 @@ The output will demonstrate FD.io VPP's trace of ping for all packets.
After examining the trace, clear it again using vpp# clear trace.
Examine arp tables
~~~~~~~~~~~~~~~~~~
-------------------
.. code-block:: console
@ -451,7 +450,7 @@ Examine arp tables
1101.5636 10.10.1.1 D e2:0f:1e:59:ec:f7 host-vpp1out
Examine routing tables
~~~~~~~~~~~~~~~~~~~~~~
-----------------------
.. code-block:: console

42
docs/guides/progressivevpp/twovppinstances.rst → docs/gettingstarted/progressivevpp/twovppinstances.rst
View File

@ -2,19 +2,19 @@
.. toctree::
Connecting two FD.io VPP Instances
----------------------------------
Connecting Two FD.io VPP Instances
==================================
.. _background-1:
memif is a very high performance, direct memory interface type which can
be used between FD.io VPP instances to form a topology. It uses a file socket
for a control channel to set up that shared memory.
be used between FD.io VPP instances. It uses a file socket for a control channel
to set up shared memory.
.. _skills-to-be-learned-1:
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
---------------------
You will learn the following new skill in this exercise:
@ -30,7 +30,7 @@ learned in previous exercises:
.. _topology-1:
Topology
^^^^^^^^
---------
.. figure:: /_images/Connecting_two_vpp_instances_with_memif.png
:alt: Connect two FD.io VPP topology
@ -40,7 +40,7 @@ Topology
.. _initial-state-1:
Initial state
^^^^^^^^^^^^^
--------------
The initial state here is presumed to be the final state from the
exercise `Create an
@ -49,27 +49,43 @@ Interface <VPP/Progressive_VPP_Tutorial#Exercise:_Create_an_Interface>`__
.. _action-running-a-second-vpp-instances-1:
Running a second FD.io VPP instances
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-------------------------------------
You should already have a FD.io VPP instance running named: vpp1.
Run a second FD.io VPP instance named: vpp2.
.. code-block:: console
$ sudo /usr/bin/vpp -c startup2.conf
....
$ sudo vppctl -s /run/vpp/cli-vpp2.sock
_______ _ _ _____ ___
__/ __/ _ \ (_)__ | | / / _ \/ _ \
_/ _// // / / / _ \ | |/ / ___/ ___/
/_/ /____(_)_/\___/ |___/_/ /_/
vpp# show version
vpp v18.07-release built by root on c469eba2a593 at Mon Jul 30 23:27:03 UTC 2018
vpp# quit
.. _action-create-memif-interface-on-vpp1-1:
Create memif interface on vpp1
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-------------------------------
Create a memif interface on vpp1:
Create a memif interface on vpp1. To connect to the instance vpp1 use the
socket **/run/vpp/cli-vpp1.sock**
.. code-block:: console
$ sudo vppctl -s /run/vpp/cli-vpp1.sock
vpp# create interface memif id 0 master
This will create an interface on vpp1 memif0/0 using /run/vpp/memif as
its socket file. The role of vpp1 for this memif inteface is 'master'.
Use your previously used skills to:
With what you have learned:
#. Set the memif0/0 state to up.
#. Assign IP address 10.10.2.1/24 to memif0/0
@ -78,7 +94,7 @@ Use your previously used skills to:
.. _action-create-memif-interface-on-vpp2-1:
Create memif interface on vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
--------------------------------
We want vpp2 to pick up the 'slave' role using the same
run/vpp/memif-vpp1vpp2 socket file
@ -99,7 +115,7 @@ Use your previously used skills to:
.. _action-ping-from-vpp1-to-vpp2-1:
Ping from vpp1 to vpp2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
------------------------
Ping 10.10.2.2 from vpp1

4
docs/gettingstarted/users/configuring/startup.rst
View File

@ -1,9 +1,5 @@
.. _startup:
.. toctree::
=======================================
VPP Configuration File - 'startup.conf'
=======================================

8
docs/gettingstarted/users/index.rst
View File

@ -1,8 +1,8 @@
.. _users:
########
Users
########
###########
For Users
###########
The Users section describe a basic VPP installation and configuration operation.
The installation and configuration of VPP can be done either manually, or by
@ -21,7 +21,7 @@ section covers the following areas:
.. toctree::
:maxdepth: 2
configutil/index.rst
installing/index.rst
configuring/index.rst
running/index.rst
configutil/index.rst

2
docs/gettingstarted/users/installing/ubuntu.rst
View File

@ -64,7 +64,7 @@ Install the mandatory packages by running the following commands:
.. code-block:: console
sudo apt-get update
sudo apt-get install vpp vpp-lib vpp-plugin
sudo apt-get install vpp-lib vpp vpp-plugins
Install the Optional Packages

1
docs/guides/index.rst
View File

@ -7,5 +7,4 @@ User Guides
.. toctree::
:maxdepth: 2
progressivevpp/index.rst
api/index.rst

176
docs/guides/progressivevpp/index.rst
View File

@ -1,176 +0,0 @@
.. _progressivevpp:
.. toctree::
########################
Progressive VPP Tutorial
########################
Overview
========
Learn to run FD.io VPP on a single Ubuntu 16.04 VM using Vagrant with this walkthrough
covering basic FD.io VPP senarios. Useful FD.io VPP commands will be used, and
will discuss basic operations, and the state of a running FD.io VPP on a system.
.. note::
This is *not* intended to be a 'How to Run in a Production Environment' set of instructions.
.. _introduction-to-vpp-vagrant:
Setting up your environment
===========================
All of these exercises are designed to be performed on an Ubuntu 16.04 (Xenial) box.
* If you have an Ubuntu 16.04 box on which you have sudo or root access, you can feel free to use that.
* If you do not, a Vagrantfile is provided to setup a basic Ubuntu 16.04 box for you in the the steps below.
Running Vagrant
===============
FD.io VPP runs in userspace. In a production environment you will often run it with DPDK to connect to real NICs or vhost to connect to VMs.
In those circumstances you usually run a single instance of FD.io VPP.
For purposes of this tutorial, it is going to be extremely useful to run multiple instances of vpp, and connect them to each other to form
a topology. Fortunately, FD.io VPP supports this.
When running multiple FD.io VPP instances, each instance needs to have specified a 'name' or 'prefix'. In the example below, the 'name' or 'prefix' is "vpp1". Note that only one instance can use the dpdk plugin, since this plugin is trying to acquire a lock on a file.
.. toctree::
settingupenvironment.rst
The DPDK Plugin will be disabled for this section. The link below demonstrates how this is done.
.. toctree::
removedpdkplugin.rst
Start a FD.io VPP shell using vppctl
====================================
The command *$ sudo vppctl* will launch a FD.io VPP shell with which you can run multiple FD.io VPP commands interactively by running:
.. code-block:: console
$ sudo vppctl
_______ _ _ _____ ___
__/ __/ _ \ (_)__ | | / / _ \/ _ \
_/ _// // / / / _ \ | |/ / ___/ ___/
/_/ /____(_)_/\___/ |___/_/ /_/
vpp# show ver
vpp v18.07-release built by root on c469eba2a593 at Mon Jul 30 23:27:03 UTC 2018
Create an Interface
===================
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
#. Create a veth interface in Linux host
#. Assign an IP address to one end of the veth interface in the Linux host
#. Create a vpp host-interface that connected to one end of a veth interface via AF_PACKET
#. Add an ip address to a vpp interface
.. toctree::
interface.rst
Traces
======
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
#. Setup a 'trace'
#. View a 'trace'
#. Clear a 'trace'
#. Verify using ping from host
#. Ping from vpp
#. Examine Arp Table
#. Examine ip fib
.. toctree::
traces.rst
Routing
=======
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
In this exercise you will learn these new skills:
#. Add route to Linux Host routing table
#. Add route to FD.io VPP routing table
And revisit the old ones:
#. Examine FD.io VPP routing table
#. Enable trace on vpp1 and vpp2
#. ping from host to FD.io VPP
#. Examine and clear trace on vpp1 and vpp2
#. ping from FD.io VPP to host
#. Examine and clear trace on vpp1 and vpp2
.. toctree::
routing.rst
Connecting Two FD.io VPP Instances
==================================
memif is a very high performance, direct memory interface type which can
be used between FD.io VPP instances to form a topology. It uses a file socket
for a control channel to set up that shared memory.
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
You will learn the following new skill in this exercise:
#. Create a memif interface between two FD.io VPP instances
You should be able to perform this exercise with the following skills
learned in previous exercises:
#. Run a second FD.io VPP instance
#. Add an ip address to a FD.io VPP interface
#. Ping from FD.io VPP
.. toctree::
twovppinstances.rst
Switching
=========
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
#. Associate an interface with a bridge domain
#. Create a loopback interaface
#. Create a BVI (Bridge Virtual Interface) for a bridge domain
#. Examine a bridge domain
.. toctree::
switching.rst
Source NAT
==========
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^
#. Abusing networks namespaces for fun and profit
#. Configuring snat address
#. Configuring snat inside and outside interfaces
.. toctree::
sourceNAT.rst

18
docs/guides/progressivevpp/removedpdkplugin.rst
View File

@ -1,18 +0,0 @@
.. _removedpdkplugin:
.. toctree::
Removing the DPDK Plugin
------------------------
For the purposes of this tutorial, the dpdk plugin will be removed.
To do this edit the *startup.conf* file with the following,
your *startup.conf* file may already have this line commented, and may just need to
uncomment it:
.. code-block:: console
plugins
{
plugin dpdk_plugin.so { disable }
}

21
docs/guides/progressivevpp/settingupenvironment.rst
View File

@ -1,21 +0,0 @@
.. _settingupenvironment:
.. toctree::
Setting up FD.io VPP environment with Vagrant
---------------------------------------------
Refer to :ref:`this guide<installingVboxVagrant>` for setting up a Virtual Box with Vagrant
After setting up Vagrant, use these commands on your Vagrant directory to boot the VM:
.. code-block:: console
# vagrant up
.. code-block:: console
# vagrant ssh
Afterwards, configure FD.io VPP on the Vagrant system following the steps on the :ref:`configutil` guide.

159
docs/guides/progressivevpp/sourceNAT.rst
View File

@ -1,159 +0,0 @@
.. _sourceNAT:
.. toctree::
Source NAT
==========
Skills to be Learned
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#. Abusing networks namespaces for fun and profit
#. Configuring snat address
#. Configuring snat inside and outside interfaces
FD.io VPP command learned in this exercise
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#. `snat add interface
address <https://docs.fd.io/vpp/17.04/clicmd_src_plugins_snat.html#clicmd_snat_add_interface_address>`__
#. `set interface
snat <https://docs.fd.io/vpp/17.04/clicmd_src_plugins_snat.html#clicmd_set_interface_snat>`__
Topology
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
.. figure:: /_images/SNAT_Topology.jpg
:alt: SNAT Topology
SNAT Topology
Initial state
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Unlike previous exercises, for this one you want to start tabula rasa.
Note: You will lose all your existing config in your FD.io VPP instances!
To clear existing config from previous exercises run:
.. code-block:: console
ps -ef | grep vpp | awk '{print $2}'| xargs sudo kill
$ sudo ip link del dev vpp1host
$ sudo ip link del dev vpp1vpp2
Install vpp-plugins
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Snat is supported by a plugin, so vpp-plugins need to be installed
.. code-block:: console
$ sudo apt-get install vpp-plugins
Create FD.io VPP instance
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Create one FD.io VPP instance named vpp1.
Confirm snat plugin is present:
.. code-block:: console
vpp# show plugins
Plugin path is: /usr/lib/vpp_plugins
Plugins loaded:
1.ioam_plugin.so
2.ila_plugin.so
3.acl_plugin.so
4.flowperpkt_plugin.so
5.snat_plugin.so
6.libsixrd_plugin.so
7.lb_plugin.so
Create veth interfaces
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#. Create a veth interface with one end named vpp1outside and the other
named vpp1outsidehost
#. Assign IP address 10.10.1.1/24 to vpp1outsidehost
#. Create a veth interface with one end named vpp1inside and the other
named vpp1insidehost
#. Assign IP address 10.10.2.1/24 to vpp1outsidehost
Because we'd like to be able to route \*via\* our vpp instance to an
interface on the same host, we are going to put vpp1insidehost into a
network namespace
Create a new network namespace 'inside'
.. code-block:: console
$ sudo ip netns add inside
Move interface vpp1inside into the 'inside' namespace:
.. code-block:: console
$ sudo ip link set dev vpp1insidehost up netns inside
Assign an ip address to vpp1insidehost
.. code-block:: console
$ sudo ip netns exec inside ip addr add 10.10.2.1/24 dev vpp1insidehost
Create a route inside the netns:
.. code-block:: console
$ sudo ip netns exec inside ip route add 10.10.1.0/24 via 10.10.2.2
Configure vpp outside interface
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#. Create a vpp host interface connected to vpp1outside
#. Assign ip address 10.10.1.2/24
#. Create a vpp host interface connected to vpp1inside
#. Assign ip address 10.10.2.2/24
Configure snat
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Configure snat to use the address of host-vpp1outside
.. code-block:: console
vpp# snat add interface address host-vpp1outside
Configure snat inside and outside interfaces
.. code-block:: console
vpp# set interface snat in host-vpp1inside out host-vpp1outside
Prepare to Observe Snat
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Observing snat in this configuration is interesting. To do so, vagrant
ssh a second time into your VM and run:
.. code-block:: console
$ sudo tcpdump -s 0 -i vpp1outsidehost
Also enable tracing on vpp1
Ping via snat
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
.. code-block:: console
$ sudo ip netns exec inside ping -c 1 10.10.1.1
Confirm snat
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Examine the tcpdump output and vpp1 trace to confirm snat occurred.

4
docs/reference/vppvagrant/index.rst
View File

@ -1,4 +1,4 @@
.. _vppcontainers:
.. _vppvagrant:
VM's with Vagrant
=================
@ -13,5 +13,3 @@ This reference guide will cover using Vagrant to boot a VM (virtual machine).
boxSetup
VagrantVMSetup
vppVagrantfile

6
docs/reference/vppvagrant/installingVboxVagrant.rst
View File

@ -1,7 +1,5 @@
.. _installingVboxVagrant:
.. toctree::
Installing Vbox and Vagrant
===========================
@ -19,6 +17,8 @@ If you're on Ubuntu, perform:
$ sudo apt-get install virtualbox
If you want to download a newer version of Virtual Box or one specific to your OS and architecture, go to the `Virtual Box download page <https://www.virtualbox.org/wiki/Downloads>`_.
Installing Vagrant
__________________
@ -35,4 +35,4 @@ This is a similar command, but on a 64-bit version of Debian:
$ sudo apt-get install https://releases.hashicorp.com/vagrant/2.1.2/vagrant_2.1.2_x86_64.deb
If you want to download a newer version of Vagrant or one specific to your OS and architecture, go to the Vagrant `download page <https://www.vagrantup.com/downloads.html>`_, right-click and copy the link address for your specified version, and replace the above install command for your respective OS and architechure.
If you want to download a newer version of Vagrant or one specific to your OS and architecture, go to the `Vagrant download page <https://www.vagrantup.com/downloads.html>`_.