FD.io (Fast data – Input/Output), the Universal Dataplane, is a collection of several projects and libraries to amplify the transformation that began with Data Plane Development Kit (DPDK) to support flexible, programmable and composable services on a generic hardware platform. FD.io offers the Software Defined Infrastructure developer community a landing site with multiple projects fostering innovations in software-based packet processing towards the creation of high-throughput, low-latency and resource-efficient IO services suitable to many architectures (x86, ARM, and PowerPC) and deployment environments (bare metal, VM, container).
A key component is the Vector Packet Processing (VPP) library donated by Cisco. This code is already running in products on the market today. The VPP library is highly modular, allowing for new graph nodes to be easily “plugged in” without changes to the underlying code base. This gives developers the potential to easily build any number of packet processing solutions with varying forwarding graphs.
Along with VPP, FD.io leverages DPDK capabilities in support of additional projects including NSH_ SFC, Honeycomb, and ONE to accelerate the NFV data planes. The FD.io projects are aligning to other key open source initiatives in support of NFV and SDN, such as OPNFV . See the Fast data Stack (FTS) project proposal within OPNFV for overall objectives; the proposal calls for an optional virtualized data plane integrated with OpenStack and OpenDaylight ( see FastDataStacks).
As the universal data plane, FD.io is helping enable the developer community achieve the highest performance, throughput, and resource efficiency in cloud networking, container networking, NFVI, bare metal networking, and SmartNic acceleration. Recently FD.io extended its lead as the fastest most scalable software data plane on earth by demonstrating terabit throughput levels with multi-million fib table entries on standard high-volume server platforms utilizing the new Intel® Xeon® Scalable Processor.
Recent testing of FD.io release 17.04 shows impressive gains in performance on Intel’s newest platform when switching and routing layer 2/3 traffic. With the prior generation Intel® Xeon® Processor E7-8890v3, FD.io testing showed aggregate forwarding rate of 480 Gbps (200 Mpps) for 4-Socket machine (using 4 of E7-8890v3 CPU configuration); however, the same FD.io tests run on two 2-Socket blades (e.g. a modern 2RU server) with the new Intel® Xeon® Platinum 8168 CPUs (using four of 8168 CPUs in two by two-socket configuration), within the same power budget, show increase of forwarding rate to 948 Gbps (400 Mpps) benefiting from the PCIe bandwidth increase of the new CPUs, and the overall decrease in cycles-per-packet due to CPU micro-architecture improvements.
For customers already building their solutions using FD.io, these types of performance gains are realized without further development effort as the software is optimized to take full advantage the additional PCIe bandwidth and other CPU capabilities as hardware become available.