diff options
| author |   John DeNisco <jdenisco@cisco.com> | 2018-07-26 12:45:10 -0400 |
|---|---|---|
| committer |   Dave Barach <openvpp@barachs.net> | 2018-07-26 18:34:47 +0000 |
| commit | 06dcd45ff81e06bc8cf40ed487c0b2652d346a5a (patch) | |
| tree | 71403f9d422c4e532b2871a66ab909bd6066b10b /docs/overview | |
| parent | 1d65279ffecd0f540288187b94cb1a6b84a7a0c6 (diff) | |
Initial commit of Sphinx docs
Change-Id: I9fca8fb98502dffc2555f9de7f507b6f006e0e77
Signed-off-by: John DeNisco <jdenisco@cisco.com>
Diffstat (limited to 'docs/overview')
26 files changed, 703 insertions, 0 deletions
diff --git a/docs/overview/features/controlplane.rst b/docs/overview/features/controlplane.rst new file mode 100644 index 00000000000..ddf0d34c69f --- /dev/null +++ b/docs/overview/features/controlplane.rst @@ -0,0 +1,10 @@ +.. _cp: + +============= +Control Plane +============= + +* DHCP client/proxy + +* DHCPv6 Proxy + diff --git a/docs/overview/features/devices.rst b/docs/overview/features/devices.rst new file mode 100644 index 00000000000..f3ecd32c825 --- /dev/null +++ b/docs/overview/features/devices.rst @@ -0,0 +1,33 @@ +.. _dev: + +======= +Devices +======= + +Hardware +-------- +* `DPDK <https://www.dpdk.org/>`_ + + * `Network Interfaces <https://doc.dpdk.org/guides/nics/>`_ + * `Cryptographic Devices <https://doc.dpdk.org/guides/cryptodevs/>`_ + +* `Open Data Plane <https://github.com/FDio/odp4vpp>`_ +* `Intel Ethernet Adaptive Virtual Function <https://www.intel.com/content/dam/www/public/us/en/documents/product-specifications/ethernet-adaptive-virtual-function-hardware-spec.pdf>`_ + +Operating System +---------------- +* `Netmap <http://info.iet.unipi.it/~luigi/netmap/>`_ +* `af_packet <http://man7.org/linux/man-pages/man7/packet.7.html>`_ +* Tap V2 (FastTap) + +Virtualization: +--------------- +* SSVM +* Vhost / VirtIO + +Containers +---------- + +* Vhost-user +* MemIF + diff --git a/docs/overview/features/index.rst b/docs/overview/features/index.rst new file mode 100644 index 00000000000..43d0503d999 --- /dev/null +++ b/docs/overview/features/index.rst @@ -0,0 +1,32 @@ +.. _features: + +======== +Features +======== + +.. rst-class:: center-align-table + ++-------------------------+-----------+-----------+ +| :ref:`sdn` | | | ++------------+------------+ :ref:`cp` | | +| | :ref:`l4` | | | +| +------------+-----------+ :ref:`pg` | +| :ref:`tun` | :ref:`l3` | | | +| +------------+ :ref:`tm` | | +| | :ref:`l2` | | | ++------------+------------+-----------+-----------+ +| :ref:`dev` | ++-------------------------------------------------+ + +.. toctree:: + :hidden: + + devices.rst + integrations.rst + trafficmanagement.rst + l2.rst + l3.rst + l4.rst + tunnels.rst + controlplane.rst + plugins.rst diff --git a/docs/overview/features/integrations.rst b/docs/overview/features/integrations.rst new file mode 100644 index 00000000000..848e8648438 --- /dev/null +++ b/docs/overview/features/integrations.rst @@ -0,0 +1,5 @@ +.. _sdn: + +======================== +SDN & Cloud Integrations +======================== diff --git a/docs/overview/features/l2.rst b/docs/overview/features/l2.rst new file mode 100644 index 00000000000..a567356209b --- /dev/null +++ b/docs/overview/features/l2.rst @@ -0,0 +1,56 @@ +.. _l2: + +======= +Layer 2 +======= + +MAC Layer +--------- +* Ethernet + +Discovery +--------- +* Cisco Discovery Protocol +* Link Layer Discovery Protocol (LLDP) + +Link Layer Control Protocol +--------------------------- +* Bit Index Explicit Replication – Link Layer Multi-cast forwarding. +* Link Layer Control (LLC) - multiplex protocols over the MAC layer. +* Spatial Reuse Protocol (SRP) +* High-Level Data Link Control (HDLC) +* Logical link control (LLC) +* Link Agg Control Protocol (Active/Active, Active/Passive) – 18.04 + +Virtual Private Networks +------------------------ +* MPLS + + * MPLS-o-Ethernet – Deep label stacks supported + +* Virtual Private LAN Service (VPLS) +* VLAN +* Q-in-Q +* Tag-rewrite (VTR) - push/pop/Translate (1:1,1:2, 2:1,2:2) +* Ethernet flow point Filtering +* Layer 2 Cross Connect + +Bridging +--------- +* Bridge Domains +* MAC Learning (50k addresses) +* Split-horizon group support +* Flooding + +ARP +--- +* Proxy +* Termination +* Bidirectional Forwarding Detection + +Integrated Routing and Bridging (IRB) +------------------------------------- +* Flexibility to both route and switch between groups of ports. +* Bridged Virtual Interface (BVI) Support, allows traffic switched traffic to be routed. + + diff --git a/docs/overview/features/l3.rst b/docs/overview/features/l3.rst new file mode 100644 index 00000000000..fd90b13efd9 --- /dev/null +++ b/docs/overview/features/l3.rst @@ -0,0 +1,55 @@ +.. _l3: + +======= +Layer 3 +======= + +IP Layer +-------- +* ICMP +* IPv4 +* IPv6 +* IPSEC +* Link Local Addressing + +MultiCast +--------- +* Multicast FiB +* IGMP + +Virtual Routing and forwarding (VRF) +------------------------------------ +* VRF scaling, thousands of tables. +* Controlled cross-VRF lookups + +Multi-path +---------- +* Equal Cost Multi Path (ECMP) +* Unequal Cost Multi Path (UCMP) + +IPv4 +---- +* ARP +* ARP Proxy +* ARP Snooping + +IPv6 +---- +* Neighbour discovery (ND) +* ND Proxy +* Router Advertisement +* Segment Routing +* Distributed Virtual Routing Resolution + +Forwarding Information Base (FIB) +--------------------------------- + +* Hierarchical FIB +* Memory efficient +* Multi-million entry scalable +* Lockless/concurrent updates +* Recursive lookups +* Next hop failure detection +* Shared FIB adjacencies +* Multicast support +* MPLS support diff --git a/docs/overview/features/l4.rst b/docs/overview/features/l4.rst new file mode 100644 index 00000000000..052c6b1b45a --- /dev/null +++ b/docs/overview/features/l4.rst @@ -0,0 +1,5 @@ +.. _l4: + +======= +Layer 4 +======= diff --git a/docs/overview/features/plugins.rst b/docs/overview/features/plugins.rst new file mode 100644 index 00000000000..99431134f7f --- /dev/null +++ b/docs/overview/features/plugins.rst @@ -0,0 +1,7 @@ +.. _pg: + +======= +Plugins +======= + +* iOAM diff --git a/docs/overview/features/trafficmanagement.rst b/docs/overview/features/trafficmanagement.rst new file mode 100644 index 00000000000..9320165ec09 --- /dev/null +++ b/docs/overview/features/trafficmanagement.rst @@ -0,0 +1,55 @@ +.. _tm: + +================== +Traffic Management +================== + +IP Layer Input Checks +--------------------- +* Source Reverse Path Forwarding +* Time To Live expiration +* IP header checksum +* Layer 2 Length < IP Length + +Classifiers +----------- +* Multiple million Classifiers - Arbitrary N-tuple + +Policers +-------- +* Colour Aware & Token Bucket +* Rounding Closest/Up/Down +* Limits in PPS/KBPS +* Types: + + * Single Rate Two Colour + * Single Rate Three Colour + * Dual Rate Three Colour + +* Action Triggers + + * Conform + * Exceed + * Violate + +* Actions Type + + * Drop + * Transmit + * Mark-and-transmit + +Switched Port Analyzer (SPAN) +* mirror traffic to another switch port + +ACLs +---- + * Stateful + * Stateless + +COP +--- + +MAC/IP Pairing +-------------- +(security feature). + diff --git a/docs/overview/features/tunnels.rst b/docs/overview/features/tunnels.rst new file mode 100644 index 00000000000..352c4393099 --- /dev/null +++ b/docs/overview/features/tunnels.rst @@ -0,0 +1,32 @@ +.. _tun: + +======= +Tunnels +======= + +Layer 2 +------- +* L2TP +* PPP +* VLAN + +Layer 3 +------- +* Mapping of Address and Port with Encapsulation (MAP-E) +* Lightweight IPv4 over IPv6 + + * An Extension to the Dual-Stack Lite Architecture + +* GENEVE +* VXLAN + +Segment Routing +--------------- +* IPv6 +* MPLS + +Generic Routing Encapsulation (GRE) +* GRE over IPSEC +* GRE over IP +* MPLS +* NSH diff --git a/docs/overview/index.rst b/docs/overview/index.rst new file mode 100644 index 00000000000..d3da0e211fa --- /dev/null +++ b/docs/overview/index.rst @@ -0,0 +1,13 @@ +.. _overview: + +========================================= +Overview +========================================= + +.. toctree:: + :maxdepth: 1 + + whatisvpp/index.rst + features/index.rst + performance/index.rst + supported.rst diff --git a/docs/overview/performance/current_ipv4_throughput.rst b/docs/overview/performance/current_ipv4_throughput.rst new file mode 100644 index 00000000000..3a34da76ca0 --- /dev/null +++ b/docs/overview/performance/current_ipv4_throughput.rst @@ -0,0 +1,12 @@ +.. _current_ipv4_throughput: + +.. toctree:: + +IPv4 Routed-Forwarding Performance Tests +**************************************** + +VPP NDR 64B packet throughput in 1t1c setup (1thread, 1core) is presented in the graph below. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/rls1804/report/_static/vpp/64B-1t1c-ethip4-ip4-ndrdisc.html" width="1200" height="1000" frameborder="0"> diff --git a/docs/overview/performance/current_ipv6_throughput.rst b/docs/overview/performance/current_ipv6_throughput.rst new file mode 100644 index 00000000000..119f28f5665 --- /dev/null +++ b/docs/overview/performance/current_ipv6_throughput.rst @@ -0,0 +1,16 @@ +.. _current_ipv6_throughput: + +.. toctree:: + +IPv6 Routed-Forwarding Performance Tests +**************************************** + +VPP NDR 78B packet throughput in 1t1c setup (1 thread, 1 core) is presented in the graph below. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/rls1801/report/_static/vpp/78B-1t1c-ethip6-ip6-ndrdisc.html" width="1200" height="1000" frameborder="0"> + + + + diff --git a/docs/overview/performance/current_l2_throughput.rst b/docs/overview/performance/current_l2_throughput.rst new file mode 100644 index 00000000000..532e6dc7163 --- /dev/null +++ b/docs/overview/performance/current_l2_throughput.rst @@ -0,0 +1,12 @@ + .. _current_l2_throughput: + +.. toctree:: + +L2 Ethernet Switching Throughput Tests +*************************************** + +VPP NDR 64B packet throughput in 1 Core, 1 Thread setup, is presented in the graph below. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/rls1801/report/_static/vpp/64B-1t1c-l2-sel2-ndrdisc.html" width="1200" height="1000" frameborder="0"> diff --git a/docs/overview/performance/current_ndr_throughput.rst b/docs/overview/performance/current_ndr_throughput.rst new file mode 100644 index 00000000000..c7348314ef6 --- /dev/null +++ b/docs/overview/performance/current_ndr_throughput.rst @@ -0,0 +1,13 @@ +.. _current_ndr_throughput: + +.. toctree:: + +NDR Performance Tests +********************* + +This is a VPP NDR 64B packet throughput in 1 Core, 1 Thread setup, live graph of the NDR (No Drop Rate) L2 Performance Tests. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/rls1804/report/_static/vpp/64B-1t1c-l2-sel1-ndrdisc.html" width="800" height="1000" frameborder="0"> + diff --git a/docs/overview/performance/index.rst b/docs/overview/performance/index.rst new file mode 100644 index 00000000000..4255d495c38 --- /dev/null +++ b/docs/overview/performance/index.rst @@ -0,0 +1,63 @@ +.. _performance: + +Performance +=========== + +Overview +^^^^^^^^ + +One of the benefits of FD.io VPP, is high performance on relatively low-power computing, this performance is based on the following features: + +* A high-performance user-space network stack designed for commodity hardware. + + - L2, L3 and L4 features and encapsulations. + +* Optimized packet interfaces supporting a multitude of use cases. + + - An integrated vhost-user backend for high speed VM-to-VM connectivity. + - An integrated memif container backend for high speed Container-to-Container connectivity. + - An integrated vhost based interface to punt packets to the Linux Kernel. + +* The same optimized code-paths run execute on the host, and inside VMs and Linux containers. +* Leverages best-of-breed open source driver technology: `DPDK <https://www.dpdk.org/>`_. +* Tested at scale; linear core scaling, tested with millions of flows and mac addresses. + +These features have been designed to take full advantage of common micro-processor optimization techniques, such as: + +* Reducing cache and TLS misses by processing packets in vectors. +* Realizing `IPC <https://en.wikipedia.org/wiki/Instructions_per_cycle>`_ gains with vector instructions such as: SSE, AVX and NEON. +* Eliminating mode switching, context switches and blocking, to always be doing useful work. +* Cache-lined aliged buffers for cache and memory efficiency. + + +Packet Throughput Graphs +^^^^^^^^^^^^^^^^^^^^^^^^ + +These are some of the packet throughput graphs for FD.io VPP 18.04 from the CSIT `18.04 benchmarking report <https://docs.fd.io/csit/rls1804/report/>`_. + +.. toctree:: + + current_l2_throughput.rst + current_ndr_throughput.rst + current_ipv4_throughput.rst + current_ipv6_throughput.rst + +Trending Throughput Graphs +^^^^^^^^^^^^^^^^^^^^^^^^^^ + +These are some of the trending packet throughput graphs from the CSIT `trending dashboard <https://docs.fd.io/csit/master/trending/introduction/index.html>`_. **Please note that**, performance in the trending graphs will change on a nightly basis in line with the software development cycle. + +.. toctree:: + + trending_l2_throughput.rst + trending_ipv4_throughput.rst + trending_ipv6_throughput.rst + +For More information on CSIT +^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +These are FD.io Continuous System Integration and Testing (CSIT)'s documentation links. + +* `CSIT Code Documentation <https://docs.fd.io/csit/master/doc/overview.html>`_ +* `CSIT Test Overview <https://docs.fd.io/csit/rls1804/report/introduction/overview.html>`_ +* `VPP Performance Dashboard <https://docs.fd.io/csit/master/trending/introduction/index.html>`_ diff --git a/docs/overview/performance/trending_ipv4_throughput.rst b/docs/overview/performance/trending_ipv4_throughput.rst new file mode 100644 index 00000000000..4ba028b9e59 --- /dev/null +++ b/docs/overview/performance/trending_ipv4_throughput.rst @@ -0,0 +1,14 @@ +.. _trending_ipv4_throughput: + +.. toctree:: + +IPv4 Routed-Forwarding Performance Tests +**************************************** + +This is a live graph of the IPv4 Routed Forwarding Switching Performance Tests. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/master/trending/_static/vpp/cpta-ip4-1t1c-x520.html" width="1200" height="1000" frameborder="0"> + + diff --git a/docs/overview/performance/trending_ipv6_throughput.rst b/docs/overview/performance/trending_ipv6_throughput.rst new file mode 100644 index 00000000000..8217b41e817 --- /dev/null +++ b/docs/overview/performance/trending_ipv6_throughput.rst @@ -0,0 +1,16 @@ +.. _trending_ipv6_throughput: + +.. toctree:: + +IPv6 Routed-Forwarding Performance Tests +**************************************** + +VPP NDR 78B packet throughput in 1t1c setup (1 thread, 1 core) is presented in the trending graph below. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/master/trending/_static/vpp/cpta-ip6-1t1c-x520-1.html" width="1200" height="1000" frameborder="0"> + + + + diff --git a/docs/overview/performance/trending_l2_throughput.rst b/docs/overview/performance/trending_l2_throughput.rst new file mode 100644 index 00000000000..59bb3bf6f1a --- /dev/null +++ b/docs/overview/performance/trending_l2_throughput.rst @@ -0,0 +1,14 @@ + .. _trending_l2_throughput: + +.. toctree:: + +L2 Ethernet Switching Performance Tests +*************************************** + +This is a live graph of the 1 Core, 1 Thread, L2 Ethernet Switching Performance Tests Test on the x520 NIC. + +.. raw:: html + + <iframe src="https://docs.fd.io/csit/master/trending/_static/vpp/cpta-l2-1t1c-x520.html" width="1200" height="1000" frameborder="0"> + + diff --git a/docs/overview/supported.rst b/docs/overview/supported.rst new file mode 100644 index 00000000000..801fdcc6966 --- /dev/null +++ b/docs/overview/supported.rst @@ -0,0 +1,27 @@ +.. _supported: + +.. toctree:: + +Architectures and Operating Systems +*********************************** + +Architectures +----------------------- + +* - The FD.io VPP platform supports: + + * - x86/64 + * - ARM + +Operating Systems and Packaging +------------------------------- + +FD.io VPP supports package installation on the following +recent LTS operating systems releases: + +* - Operating Systems: + + * - Debian + * - Ubuntu + * - CentOS + * - OpenSUSE diff --git a/docs/overview/whatisvpp/dataplane.rst b/docs/overview/whatisvpp/dataplane.rst new file mode 100644 index 00000000000..9113038fc57 --- /dev/null +++ b/docs/overview/whatisvpp/dataplane.rst @@ -0,0 +1,34 @@ +.. _packet-processing: + +================= +Packet Processing +================= + +* Layer 2 - 4 Network Stack + + * Fast lookup tables for routes, bridge entries + * Arbitrary n-tuple classifiers + * Control Plane, Traffic Management and Overlays + +* `Linux <https://en.wikipedia.org/wiki/Linux>`_ and `FreeBSD <https://en.wikipedia.org/wiki/FreeBSD>`_ support + + * Wide support for standard Operating System Interfaces such as AF_Packet, Tun/Tap & Netmap. + +* Wide network and cryptograhic hardware support with `DPDK <https://www.dpdk.org/>`_. +* Container and Virtualization support + + * Para-virtualized intefaces; Vhost and Virtio + * Network Adapters over PCI passthrough + * Native container interfaces; MemIF + +* Universal Data Plane: one code base, for many use cases + + * Discrete appliances; such as `Routers <https://en.wikipedia.org/wiki/Router_(computing)>`_ and `Switches <https://en.wikipedia.org/wiki/Network_switch>`_. + * `Cloud Infrastructure and Virtual Network Functions <https://en.wikipedia.org/wiki/Network_function_virtualization>`_ + * `Cloud Native Infrastructure <https://www.cncf.io/>`_ + * The same binary package for all use cases. + +* Out of the box production quality, with thanks to `CSIT <https://wiki.fd.io/view/CSIT#Start_Here>`_. + +For more information, please see :ref:`features` for the complete list. + diff --git a/docs/overview/whatisvpp/developer.rst b/docs/overview/whatisvpp/developer.rst new file mode 100644 index 00000000000..a0bb2d451c0 --- /dev/null +++ b/docs/overview/whatisvpp/developer.rst @@ -0,0 +1,24 @@ +.. _developer-friendly: + +================== +Developer Friendly +================== + +* Extensive runtime counters; throughput, `intructions per cycle <https://en.wikipedia.org/wiki/Instructions_per_cycle>`_, errors, events etc. +* Integrated pipeline tracing facilities +* Multi-language API bindings +* Integrated command line for debugging +* Fault-tolerant and upgradable + + * Runs as a standard user-space process for fault tolerance, software crashes seldom require more than a process restart. + * Improved fault-tolerance and upgradability when compared to running similar packet processing in the kernel, software updates never require system reboots. + * Development expierence is easier compared to similar kernel code + * Hardware isolation and protection (`iommu <https://en.wikipedia.org/wiki/Input%E2%80%93output_memory_management_unit>`_) + +* Built for security + + * Extensive white-box testing + * Image segment base address randomization + * Shared-memory segment base address randomization + * Stack bounds checking + * Static analysis with `Coverity <https://en.wikipedia.org/wiki/Coverity>`_ diff --git a/docs/overview/whatisvpp/extensible.rst b/docs/overview/whatisvpp/extensible.rst new file mode 100644 index 00000000000..c271dad7d14 --- /dev/null +++ b/docs/overview/whatisvpp/extensible.rst @@ -0,0 +1,39 @@ +.. _extensible: + +============================= +Extensible and Modular Design +============================= + +* Pluggable, easy to understand & extend +* Mature graph node architecture +* Full control to reorganize the pipeline +* Fast, plugins are equal citizens + +**Modular, Flexible, and Extensible** + +The FD.io VPP packet processing pipeline is decomposed into a ‘packet processing +graph’. This modular approach means that anyone can ‘plugin’ new graph +nodes. This makes VPP easily exensible and means that plugins can be +customized for specific purposes. VPP is also configurable through it's +Low-Level API. + +.. figure:: /_images/VPP_custom_application_packet_processing_graph.280.jpg + :alt: Extensible, modular graph node architecture? + + Extensible and modular graph node architecture. + +At runtime, the FD.io VPP platform assembles a vector of packets from RX rings, +typically up to 256 packets in a single vector. The packet processing graph is +then applied, node by node (including plugins) to the entire packet vector. The +received packets typically traverse the packet processing graph nodes in the +vector, when the network processing represented by each graph node is applied to +each packet in turn. Graph nodes are small and modular, and loosely +coupled. This makes it easy to introduce new graph nodes and rewire existing +graph nodes. + +Plugins are `shared libraries <https://en.wikipedia.org/wiki/Library_(computing)>`_ +and are loaded at runtime by VPP. VPP find plugins by searching the plugin path +for libraries, and then dynamically loads each one in turn on startup. +A plugin can introduce new graph nodes or rearrange the packet processing graph. +You can build a plugin completely independently of the FD.io VPP source tree, +which means you can treat it as an independent component. diff --git a/docs/overview/whatisvpp/fast.rst b/docs/overview/whatisvpp/fast.rst new file mode 100644 index 00000000000..b04c12fbe7b --- /dev/null +++ b/docs/overview/whatisvpp/fast.rst @@ -0,0 +1,16 @@ +.. _fast: + +================================ +Fast, Scalable and Deterministic +================================ + +* `Continuous integration and system testing <https://wiki.fd.io/view/CSIT#Start_Here>`_ + + * Including continuous & extensive, latency and throughput testing + +* Layer 2 Cross Connect (L2XC), typically achieve 15+ Mpps per core. +* Tested to achieve **zero** packet drops and ~15µs latency. +* Performance scales linearly with core/thread count +* Supporting millions of concurrent lookup tables entries + +Please see :ref:`performance` for more information. diff --git a/docs/overview/whatisvpp/index.rst b/docs/overview/whatisvpp/index.rst new file mode 100644 index 00000000000..f8cb25dfd1e --- /dev/null +++ b/docs/overview/whatisvpp/index.rst @@ -0,0 +1,27 @@ +.. _whatisvpp: + +========================================= +What is VPP? +========================================= + +FD.io's Vector Packet Processing (VPP) technology is a :ref:`fast`, +:ref:`packet-processing` stack that runs on commodity CPUs. It provides +out-of-the-box production quality switch/router functionality and much, much +more. FD.io VPP is at the same time, an :ref:`extensible` and +:ref:`developer-friendly` framework, capable of boot-strapping the development +of packet-processing applications. The benefits of FD.io VPP are its high +performance, proven technology, its modularity and flexibility, integrations and +rich feature set. + +FD.io VPP is vector packet processing software, to learn more about what that +means, see the :ref:`what-is-vector-packet-processing` section. + +For more detailed information on FD.io features, see the following sections: + +.. toctree:: + :maxdepth: 1 + + dataplane.rst + fast.rst + developer.rst + extensible.rst diff --git a/docs/overview/whatisvpp/what-is-vector-packet-processing.rst b/docs/overview/whatisvpp/what-is-vector-packet-processing.rst new file mode 100644 index 00000000000..994318e81c5 --- /dev/null +++ b/docs/overview/whatisvpp/what-is-vector-packet-processing.rst @@ -0,0 +1,73 @@ +:orphan: + +.. _what-is-vector-packet-processing: + +================================= +What is vector packet processing? +================================= + +FD.io VPP is developed using vector packet processing concepts, as opposed to +scalar packet processing, these concepts are explained in the following sections. + +Vector packet processing is a common approach among high performance `Userspace +<https://en.wikipedia.org/wiki/User_space>`_ packet processing applications such +as developed with FD.io VPP and `DPDK +<https://en.wikipedia.org/wiki/Data_Plane_Development_Kit>`_. The scalar based +aproach tends to be favoured by Operating System `Kernel +<https://en.wikipedia.org/wiki/Kernel_(operating_system)>`_ Network Stacks and +Userspace stacks that don't have strict performance requirements. + +**Scalar Packet Processing** + +A scalar packet processing network stack typically processes one packet at a +time: an interrupt handling function takes a single packet from a Network +Inteface, and processes it through a set of functions: fooA calls fooB calls +fooC and so on. + +.. code-block:: none + + +---> fooA(packet1) +---> fooB(packet1) +---> fooC(packet1) + +---> fooA(packet2) +---> fooB(packet2) +---> fooC(packet2) + ... + +---> fooA(packet3) +---> fooB(packet3) +---> fooC(packet3) + + +Scalar packet processing is simple, but inefficent in these ways: + +* When the code path length exceeds the size of the Microprocessor's instruction + cache (I-cache), `thrashing + <https://en.wikipedia.org/wiki/Thrashing_(computer_science)>`_ occurs as the + Microprocessor is continually loading new instructions. In this model, each + packet incurs an identical set of I-cache misses. +* The associated deep call stack will also add load-store-unit pressure as + stack-locals fall out of the Microprocessor's Layer 1 Data Cache (D-cache). + +**Vector Packet Processing** + +In contrast, a vector packet processing network stack processes multiple packets +at a time, called 'vectors of packets' or simply a 'vector'. An interrupt +handling function takes the vector of packets from a Network Inteface, and +processes the vector through a set of functions: fooA calls fooB calls fooC and +so on. + +.. code-block:: none + + +---> fooA([packet1, +---> fooB([packet1, +---> fooC([packet1, +---> + packet2, packet2, packet2, + ... ... ... + packet256]) packet256]) packet256]) + +This approach fixes: + +* The I-cache thrashing problem described above, by ammoritizing the cost of + I-cache loads across multiple packets. + +* The ineffeciences associated with the deep call stack by recieving vectors + of up to 256 packets at a time from the Network Interface, and processes them + using a directed graph of node. The graph scheduler invokes one node dispatch + function at a time, restricting stack depth to a few stack frames. + +The further optimizations that this approaches enables are pipelining and +prefetching to minimize read latency on table data and parallelize packet loads +needed to process packets. + |