summaryrefslogtreecommitdiffstats
path: root/docs/usecases/vhost
diff options
context:
space:
mode:
authorJohn DeNisco <jdenisco@cisco.com>2018-07-26 12:45:10 -0400
committerDave Barach <openvpp@barachs.net>2018-07-26 18:34:47 +0000
commit06dcd45ff81e06bc8cf40ed487c0b2652d346a5a (patch)
tree71403f9d422c4e532b2871a66ab909bd6066b10b /docs/usecases/vhost
parent1d65279ffecd0f540288187b94cb1a6b84a7a0c6 (diff)
Initial commit of Sphinx docs
Change-Id: I9fca8fb98502dffc2555f9de7f507b6f006e0e77 Signed-off-by: John DeNisco <jdenisco@cisco.com>
Diffstat (limited to 'docs/usecases/vhost')
-rw-r--r--docs/usecases/vhost/index.rst17
-rw-r--r--docs/usecases/vhost/iperf-vm.xml106
-rw-r--r--docs/usecases/vhost/vhost.rst115
-rw-r--r--docs/usecases/vhost/vhost02.rst109
-rw-r--r--docs/usecases/vhost/vhost03.rst88
-rw-r--r--docs/usecases/vhost/vhost04.rst43
-rw-r--r--docs/usecases/vhost/vhost05.rst25
-rw-r--r--docs/usecases/vhost/xmlexample.rst11
8 files changed, 514 insertions, 0 deletions
diff --git a/docs/usecases/vhost/index.rst b/docs/usecases/vhost/index.rst
new file mode 100644
index 00000000000..002ebc17639
--- /dev/null
+++ b/docs/usecases/vhost/index.rst
@@ -0,0 +1,17 @@
+.. _vhost:
+
+FD.io VPP with Virtual Machines
+===============================
+This chapter will describe how to use FD.io VPP with virtual machines. We describe
+how to create Vhost port with VPP and how to connect them to VPP. We will also discuss
+some limitations of Vhost.
+
+.. toctree::
+
+ vhost
+ vhost02
+ vhost03
+ vhost04
+ vhost05
+ xmlexample
+
diff --git a/docs/usecases/vhost/iperf-vm.xml b/docs/usecases/vhost/iperf-vm.xml
new file mode 100644
index 00000000000..be354c5f977
--- /dev/null
+++ b/docs/usecases/vhost/iperf-vm.xml
@@ -0,0 +1,106 @@
+<domain type='kvm' id='54'>
+ <name>iperf-server</name>
+ <memory unit='KiB'>1048576</memory>
+ <currentMemory unit='KiB'>1048576</currentMemory>
+ <memoryBacking>
+ <hugepages>
+ <page size='2048' unit='KiB'/>
+ </hugepages>
+ </memoryBacking>
+ <vcpu placement='static'>1</vcpu>
+ <resource>
+ <partition>/machine</partition>
+ </resource>
+ <os>
+ <type arch='x86_64' machine='pc-i440fx-xenial'>hvm</type>
+ <boot dev='hd'/>
+ </os>
+ <features>
+ <acpi/>
+ <apic/>
+ </features>
+ <cpu mode='host-model'>
+ <model fallback='allow'></model>
+ <numa>
+ <cell id='0' cpus='0' memory='262144' unit='KiB' memAccess='shared'/>
+ </numa>
+ </cpu>
+ <clock offset='utc'>
+ <timer name='rtc' tickpolicy='catchup'/>
+ <timer name='pit' tickpolicy='delay'/>
+ <timer name='hpet' present='no'/>
+ </clock>
+ <on_poweroff>destroy</on_poweroff>
+ <on_reboot>restart</on_reboot>
+ <on_crash>restart</on_crash>
+ <pm>
+ <suspend-to-mem enabled='no'/>
+ <suspend-to-disk enabled='no'/>
+ </pm>
+ <devices>
+ <emulator>/usr/bin/kvm</emulator>
+ <disk type='file' device='disk'>
+ <driver name='qemu' type='qcow2'/>
+ <source file='/tmp/xenial-mod.img'/>
+ <backingStore/>
+ <target dev='vda' bus='virtio'/>
+ <alias name='virtio-disk0'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x07' function='0x0'/>
+ </disk>
+ <disk type='file' device='cdrom'>
+ <driver name='qemu' type='raw'/>
+ <source file='/scratch/jdenisco/sae/configs/cloud-config.iso'/>
+ <backingStore/>
+ <target dev='hda' bus='ide'/>
+ <readonly/>
+ <alias name='ide0-0-0'/>
+ <address type='drive' controller='0' bus='0' target='0' unit='0'/>
+ </disk>
+ <controller type='usb' index='0' model='ich9-ehci1'>
+ <alias name='usb'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x7'/>
+ </controller>
+ <controller type='pci' index='0' model='pci-root'>
+ <alias name='pci.0'/>
+ </controller>
+ <controller type='ide' index='0'>
+ <alias name='ide'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/>
+ </controller>
+ <controller type='virtio-serial' index='0'>
+ <alias name='virtio-serial0'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/>
+ </controller>
+ <interface type='vhostuser'>
+ <mac address='52:54:00:4c:47:f2'/>
+ <source type='unix' path='/tmp//vm00.sock' mode='server'/>
+ <model type='virtio'/>
+ <alias name='net1'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/>
+ </interface>
+ <serial type='pty'>
+ <source path='/dev/pts/2'/>
+ <target port='0'/>
+ <alias name='serial0'/>
+ </serial>
+ <console type='pty' tty='/dev/pts/2'>
+ <source path='/dev/pts/2'/>
+ <target type='serial' port='0'/>
+ <alias name='serial0'/>
+ </console>
+ <input type='mouse' bus='ps2'/>
+ <input type='keyboard' bus='ps2'/>
+ <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1'>
+ <listen type='address' address='127.0.0.1'/>
+ </graphics>
+ <memballoon model='virtio'>
+ <alias name='balloon0'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x08' function='0x0'/>
+ </memballoon>
+ </devices>
+ <seclabel type='dynamic' model='apparmor' relabel='yes'>
+ <label>libvirt-2c4c9317-c7a5-4b37-b789-386ccda7348a</label>
+ <imagelabel>libvirt-2c4c9317-c7a5-4b37-b789-386ccda7348a</imagelabel>
+ </seclabel>
+</domain>
+
diff --git a/docs/usecases/vhost/vhost.rst b/docs/usecases/vhost/vhost.rst
new file mode 100644
index 00000000000..f62faade306
--- /dev/null
+++ b/docs/usecases/vhost/vhost.rst
@@ -0,0 +1,115 @@
+.. toctree::
+
+.. _vhost01:
+
+Prerequisites
+-------------
+
+For this use case we will assume FD.io VPP is installed. We will also assume the user can create and start
+basic virtual machines. This use case will use the linux virsh commands. For more information on virsh
+refer to `virsh man page <https://linux.die.net/man/1/virsh>`_.
+
+The image that we use is based on an Ubuntu cloud image downloaded from:
+`Ubuntu Cloud Images <https://cloud-images.ubuntu.com/xenial/current>`_.
+
+All FD.io VPP commands are being run from a su shell.
+
+.. _vhosttopo:
+
+Topology
+---------
+
+In this case we will use 2 systems. One system we will be running standard linux, the other will
+be running FD.io VPP.
+
+.. figure:: /_images/vhost-topo.png
+ :alt:
+
+ Vhost Use Case Topology
+
+Creating The Virtual Interface
+------------------------------
+
+We will start on the system running FD.io VPP and show that no Virtual interfaces have been created.
+We do this using the :ref:`showintcommand` command.
+
+Notice we do not have any virtual interfaces. We do have an interface (TenGigabitEthernet86/0/0) that
+is up. This interface is connected to a system running, in our example standard linux. We will use
+this system to verify our connectivity to our VM with ping.
+
+.. code-block:: console
+
+ $ sudo bash
+ # vppctl
+ _______ _ _ _____ ___
+ __/ __/ _ \ (_)__ | | / / _ \/ _ \
+ _/ _// // / / / _ \ | |/ / ___/ ___/
+ /_/ /____(_)_/\___/ |___/_/ /_/
+
+ vpp# clear interfaces
+ vpp# show int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 up
+ TenGigabitEthernet86/0/1 2 down
+ local0 0 down
+ vpp#
+
+For more information on the interface commands refer to: :ref:`intcommands`
+
+The next step will be to create the virtual port using the :ref:`createvhostuser` command.
+This command will create the virtual port in VPP and create a linux socket that the VM will
+use to connect to VPP.
+
+The port can be created using VPP as the socket server or client.
+
+Creating the VPP port:
+
+.. code-block:: console
+
+ vpp# create vhost socket /tmp/vm00.sock
+ VirtualEthernet0/0/0
+ vpp# show int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 up
+ TenGigabitEthernet86/0/1 2 down
+ VirtualEthernet0/0/0 3 down
+ local0 0 down
+ vpp#
+
+Notice the interface **VirtualEthernet0/0/0**. In this example we created the virtual interface as
+a client.
+
+We can get more detail on the vhost connection with the :ref:`showvhost` command.
+
+.. code-block:: console
+
+ vpp# show vhost
+ Virtio vhost-user interfaces
+ Global:
+ coalesce frames 32 time 1e-3
+ number of rx virtqueues in interrupt mode: 0
+ Interface: VirtualEthernet0/0/0 (ifindex 3)
+ virtio_net_hdr_sz 12
+ features mask (0xffffffffffffffff):
+ features (0x58208000):
+ VIRTIO_NET_F_MRG_RXBUF (15)
+ VIRTIO_NET_F_GUEST_ANNOUNCE (21)
+ VIRTIO_F_ANY_LAYOUT (27)
+ VIRTIO_F_INDIRECT_DESC (28)
+ VHOST_USER_F_PROTOCOL_FEATURES (30)
+ protocol features (0x3)
+ VHOST_USER_PROTOCOL_F_MQ (0)
+ VHOST_USER_PROTOCOL_F_LOG_SHMFD (1)
+
+ socket filename /tmp/vm00.sock type client errno "No such file or directory"
+
+ rx placement:
+ tx placement: spin-lock
+ thread 0 on vring 0
+ thread 1 on vring 0
+
+ Memory regions (total 0)
+
+Notice **No such file or directory** and **Memory regions (total 0)**. This is because the
+VM has not been created yet.
+
diff --git a/docs/usecases/vhost/vhost02.rst b/docs/usecases/vhost/vhost02.rst
new file mode 100644
index 00000000000..b9d1f5696c9
--- /dev/null
+++ b/docs/usecases/vhost/vhost02.rst
@@ -0,0 +1,109 @@
+.. _vhost02:
+
+Creating the Virtual Machine
+----------------------------
+
+We will now create the virtual machine. We use the "virsh create command". For the complete file we
+use refer to :ref:`xmlexample`.
+
+It is important to note that in the XML file we specify the socket path that is used to connect to
+FD.io VPP.
+
+This is done with a section that looks like this
+
+.. code-block:: console
+
+ <interface type='vhostuser'>
+ <mac address='52:54:00:4c:47:f2'/>
+ <source type='unix' path='/tmp//vm00.sock' mode='server'/>
+ <model type='virtio'/>
+ <alias name='net1'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/>
+ </interface>
+
+Notice the **interface type** and the **path** to the socket.
+
+Now we create the VM. The virsh list command shows the VMs that have been created. We start with no VMs.
+
+.. code-block:: console
+
+ $ virsh list
+ Id Name State
+ ----------------------------------------------------
+
+Create the VM with the virsh create command specifying our xml file.
+
+.. code-block:: console
+
+ $ virsh create ./iperf3-vm.xml
+ Domain iperf-server3 created from ./iperf3-vm.xml
+
+ $ virsh list
+ Id Name State
+ ----------------------------------------------------
+ 65 iperf-server3 running
+
+The VM is now created.
+
+.. note::
+
+ After a VM is created an xml file can created with "virsh dumpxml".
+
+.. code-block:: console
+
+ $ virsh dumpxml iperf-server3
+ <domain type='kvm' id='65'>
+ <name>iperf-server3</name>
+ <uuid>e23d37c1-10c3-4a6e-ae99-f315a4165641</uuid>
+ <memory unit='KiB'>262144</memory>
+ .....
+
+Once the virtual machine is created notice the socket filename shows **Success** and
+there are **Memory Regions**. At this point the VM and FD.io VPP are connected. Also
+notice **qsz 256**. This system is running an older version of qemu. A queue size of 256
+will affect vhost throughput. The qsz should be 1024. On the web you should be able to
+find ways to install a newer version of qemu or change the queue size.
+
+.. code-block:: console
+
+ vpp# show vhost
+ Virtio vhost-user interfaces
+ Global:
+ coalesce frames 32 time 1e-3
+ number of rx virtqueues in interrupt mode: 0
+ Interface: VirtualEthernet0/0/0 (ifindex 3)
+ virtio_net_hdr_sz 12
+ features mask (0xffffffffffffffff):
+ features (0x58208000):
+ VIRTIO_NET_F_MRG_RXBUF (15)
+ VIRTIO_NET_F_GUEST_ANNOUNCE (21)
+ VIRTIO_F_ANY_LAYOUT (27)
+ VIRTIO_F_INDIRECT_DESC (28)
+ VHOST_USER_F_PROTOCOL_FEATURES (30)
+ protocol features (0x3)
+ VHOST_USER_PROTOCOL_F_MQ (0)
+ VHOST_USER_PROTOCOL_F_LOG_SHMFD (1)
+
+ socket filename /tmp/vm00.sock type client errno "Success"
+
+ rx placement:
+ thread 1 on vring 1, polling
+ tx placement: spin-lock
+ thread 0 on vring 0
+ thread 1 on vring 0
+
+ Memory regions (total 2)
+ region fd guest_phys_addr memory_size userspace_addr mmap_offset mmap_addr
+ ====== ===== ================== ================== ================== ================== =============== ===
+ 0 31 0x0000000000000000 0x00000000000a0000 0x00007f1db9c00000 0x0000000000000000 0x00007f7db0400 000
+ 1 32 0x00000000000c0000 0x000000000ff40000 0x00007f1db9cc0000 0x00000000000c0000 0x00007f7d94ec0 000
+
+ Virtqueue 0 (TX)
+ qsz 256 last_avail_idx 0 last_used_idx 0
+ avail.flags 0 avail.idx 256 used.flags 1 used.idx 0
+ kickfd 33 callfd 34 errfd -1
+
+ Virtqueue 1 (RX)
+ qsz 256 last_avail_idx 8 last_used_idx 8
+ avail.flags 0 avail.idx 8 used.flags 1 used.idx 8
+ kickfd 29 callfd 35 errfd -1
diff --git a/docs/usecases/vhost/vhost03.rst b/docs/usecases/vhost/vhost03.rst
new file mode 100644
index 00000000000..ed583349bc6
--- /dev/null
+++ b/docs/usecases/vhost/vhost03.rst
@@ -0,0 +1,88 @@
+.. _vhost03:
+
+Bridge the Interfaces
+---------------------
+
+To connect the 2 interfaces we put them on an L2 bridge.
+
+Use the "set interface l2 bridge" command.
+
+.. code-block:: console
+
+ vpp# set interface l2 bridge VirtualEthernet0/0/0 100
+ vpp# set interface l2 bridge TenGigabitEthernet86/0/0 100
+ vpp# show bridge
+ BD-ID Index BSN Age(min) Learning U-Forwrd UU-Flood Flooding ARP-Term BVI-Intf
+ 100 1 0 off on on on on off N/A
+ vpp# show bridge 100 det
+ BD-ID Index BSN Age(min) Learning U-Forwrd UU-Flood Flooding ARP-Term BVI-Intf
+ 100 1 0 off on on on on off N/A
+
+ Interface If-idx ISN SHG BVI TxFlood VLAN-Tag-Rewrite
+ VirtualEthernet0/0/0 3 1 0 - * none
+ TenGigabitEthernet86/0/0 1 1 0 - * none
+ vpp# show vhost
+
+Bring the Interfaces Up
+-----------------------
+
+We can now bring all the pertinent interfaces up. We can then we will then be able to communicate
+with the VM from the remote system running Linux.
+
+Bring the interfaces up with :ref:`setintstate` command.
+
+.. code-block:: console
+
+ vpp# set interface state VirtualEthernet0/0/0 up
+ vpp# set interface state TenGigabitEthernet86/0/0 up
+ vpp# sh int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 up rx packets 2
+ rx bytes 180
+ TenGigabitEthernet86/0/1 2 down
+ VirtualEthernet0/0/0 3 up tx packets 2
+ tx bytes 180
+ local0 0 down
+
+Ping from the VM
+----------------
+
+The remote Linux system has an ip address of "10.0.0.2" we can now reach it from the VM.
+
+Use the "virsh console" command to attach to the VM. "ctrl-D" to exit.
+
+.. code-block:: console
+
+ $ virsh console iperf-server3
+ Connected to domain iperf-server3
+ Escape character is ^]
+
+ Ubuntu 16.04.3 LTS iperfvm ttyS0
+ .....
+
+ root@iperfvm:~# ping 10.0.0.2
+ 64 bytes from 10.0.0.2: icmp_seq=1 ttl=64 time=0.285 ms
+ 64 bytes from 10.0.0.2: icmp_seq=2 ttl=64 time=0.154 ms
+ 64 bytes from 10.0.0.2: icmp_seq=3 ttl=64 time=0.159 ms
+ 64 bytes from 10.0.0.2: icmp_seq=4 ttl=64 time=0.208 ms
+
+
+On VPP you can now see the packet counts increasing. The packets from the VM are seen as **rx packets**
+on **VirtualEthernet0/0/0**, they are then bridged to **TenGigabitEthernet86/0/0** and are seen leaving the
+system as **tx packets**. The reverse is true on the way in.
+
+.. code-block:: console
+
+ vpp# sh int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 up rx packets 16
+ rx bytes 1476
+ tx packets 14
+ tx bytes 1260
+ TenGigabitEthernet86/0/1 2 down
+ VirtualEthernet0/0/0 3 up rx packets 14
+ rx bytes 1260
+ tx packets 16
+ tx bytes 1476
+ local0 0 down
+ vpp#
diff --git a/docs/usecases/vhost/vhost04.rst b/docs/usecases/vhost/vhost04.rst
new file mode 100644
index 00000000000..256c0b8ffa4
--- /dev/null
+++ b/docs/usecases/vhost/vhost04.rst
@@ -0,0 +1,43 @@
+.. _vhost04:
+
+Cleanup
+-------
+
+Destroy the VMs with "virsh destroy"
+
+.. code-block:: console
+
+ cto@tf-ucs-3:~$ virsh list
+ Id Name State
+ ----------------------------------------------------
+ 65 iperf-server3 running
+
+ cto@tf-ucs-3:~$ virsh destroy iperf-server3
+ Domain iperf-server3 destroyed
+
+
+Delete the Virtual port in FD.io VPP
+
+.. code-block:: console
+
+ vpp# delete vhost-user VirtualEthernet0/0/0
+ vpp# show int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 up rx packets 21
+ rx bytes 1928
+ tx packets 19
+ tx bytes 1694
+ TenGigabitEthernet86/0/1 2 down
+ local0 0 down
+
+Restart FD.io VPP
+
+.. code-block:: console
+
+ # service vpp restart
+ # vppctl show int
+ Name Idx State Counter Count
+ TenGigabitEthernet86/0/0 1 down
+ TenGigabitEthernet86/0/1 2 down
+ local0 0 down
+
diff --git a/docs/usecases/vhost/vhost05.rst b/docs/usecases/vhost/vhost05.rst
new file mode 100644
index 00000000000..4eba6e17101
--- /dev/null
+++ b/docs/usecases/vhost/vhost05.rst
@@ -0,0 +1,25 @@
+.. _vhost05:
+
+Limitations
+-----------
+There are some limitations when using the qemu vhost driver. Some are described in this section.
+
+Performance
+^^^^^^^^^^^
+
+VPP performance with vHost is limited by the Qemu vHost driver. FD.io VPP 18.04 CSIT vHost testing
+shows with 2 threads, 2 cores and a Queue size of 1024 the maximum NDR throughput was about 7.5 Mpps.
+This is about the limit at this time.
+
+For all the details on the CSIT VM vhost connection refer to the
+`CSIT VM vHost performance tests <https://docs.fd.io/csit/rls1804/report/vpp_performance_tests/packet_throughput_graphs/vm_vhost.html>`_.
+
+
+Features
+^^^^^^^^
+
+These are the features not supported with FD.io VPP vHost.
+
+* VPP implements vHost in device mode only. VPP is intended to work with Qemu which implements vHost in driver mode, it does not implement vHost driver mode.
+* VPP vHost implementation does not support checksum or transmit segmentation offload.
+* VPP vHost implementation does not support packet receive filtering feature for controlling receive traffic.
diff --git a/docs/usecases/vhost/xmlexample.rst b/docs/usecases/vhost/xmlexample.rst
new file mode 100644
index 00000000000..34445da0ccb
--- /dev/null
+++ b/docs/usecases/vhost/xmlexample.rst
@@ -0,0 +1,11 @@
+.. _xmlexample01:
+
+The XML File
+------------
+
+An example of a file that could be used with the virsh create command.
+
+.. literalinclude:: iperf-vm.xml
+ :language: XML
+ :emphasize-lines: 42-49, 74-80
+