Ray Kinsella

The Userspace networking ecosystems has really matured since DPDK (the Data Plane Development Kit) was first open sourced in March 2013. Since then, a whole ecosystem of technologies developed with it or seeking to becoming a better DPDK has emerged.

Sometimes controversially data plane networking (the packet processing) which was once the sole responsibility of the Operating System Kernel is now being handled entirely in Userspace, by applications built with tools such as DPDK.

This talks discusses:-

1. Why the phenomenon exists? Where did it come from? Why are people eager to re-invent the wheel?
2. A (very) brief survey of user space data plane networking projects.
3. The case for better inter-operability, need the Kernel and Userspace data plane really be two separate Islands?
4. How to collaborate to achieve a better future?

Please note that this talk was originally scheduled to be given on Sunday at 14:00. The talk originally in this slot, Theo: The Authorized Keys Manager by Michele Azzolari will now take place on Sunday at 14:00.

Numerous open-source projects develop software-defined data planes targeting network use cases such as Discrete Appliances, Cloud Infrastructure, Virtual Network Functions and now also Cloud-Native deployments. These may be based on foundation toolkits such DPDK, eBPF/XDP, Snabb and so on, and may implement a diverse range of network functions, applied in many combinations and on different compute platforms and devices.

The need for a consistent and repeatable, use-case driven performance validation and benchmarking approach has never been greater. Enabling both the development community and the end-user to understand, measure and verify expected performance.

Achieving great performance with Network Software, should not be accidental, it should be data-driven!

This talk explains how the FD.io CSIT project aims to meet this need by developing and providing Continuous Performance Lab platform for benchmarking, validation, performance trending and regression detection. Founded on multi-vendor collaboration, CSIT leverages deep multi-platform understanding and telemetry tools to analyze and correlate benchmarking results leading to consistent, repeatable and reliable performance validation the FD.io user base can rely on.

This talk will cover

• A brief overview of FD.io CSIT
• Recent FD.io VPP release performance data is presented and discussed in this context.
• Recent CSIT Community Innovations Performance Trending graphs, MLRSearch and PMU Tooling are presented.
• Debugging a real system setup issue with FD.io CSIT is also covered

Short presentation on hacking the Oregon Scientific WeatherStation with a proprietary interface. To free the data within.

Short presentation on hacking an Oregon Scientific WeatherStation with a proprietary interface. The Weather Station itself was provided with a USB port, with some clunky Microsoft Windows Software to dump the data to CSV. Pulling the data directly out of the Weather Station by the port provided was not going to be an option.

Instead Open Source software came to the rescue; in the form of a Arduino Sketch implementing Manchester Encoding with a €2 RXB6 module. With this, I was able to intercept the 433 Mz Sensor Data and upload to the internet with WeeWX.

This is a story of a €12 investment and Open Source, opening a World of Weather Data.

Embracing containers and micro-services with their scaling model and resiliency is the new frontier in software packet processing. This talk will describe the reasons, the challenges and opportunities of this trend.

DPDK revolutionized software packet processing initially for discrete appliances and then for Virtual Network Functions. Containers and micro-services technologies are extensively used as a means to scale up and out in the Cloud.

Containers and micro-services now include Comms Service Providers among their advocates, and embracing these technologies with their scaling model and resiliency is the new frontier in software packet processing. However there are challenges in decomposing Software Network Functions into Data-plane Micro-Services.

This talk will describe:- 1. The challenges of enabling robustness and scaling, while still meeting data-plane performance goals. 2. The use cases, the trends and the timelines. 3. The enabling work to date in DPDK and FD.io VPP, and the next steps.

The Transport Layer Development Kit (TLDK) is new Layer 4 implementation, developed for DPDK (dpdk.org) and VPP (https://wiki.fd.io/view/VPP).

TCP and UDP socket based applications are core technologies in the modern data-center. The introduction of Orchestration, SDN controllers, vSwitches and Virtual Dataplane have seen a huge change in how network packets are delivered to data-centre nodes. While the stacks that consume these packets, the socket based applications have evolved much more slowly. Virtual switches like VPP and OVS-DPDK, have benefited from the vector packet processing technology in DPDK to scale process millions of packets per second, while TCP and UDP socket based applications have not yet benefited. There have been a number recent efforts to port various L4 implementations to DPDK, and while these usually give a modest performance benefit the performance of these is not yet revolutionary.

TLDK has been developed from the ground up to be a very fast and scalable UDP and TCP implementation, targeted for use with DPDK and VPP. In this presentation will provide an overview of TLDK design and implementation. This deep-dive will include:-

• Overview of the TLDK UDP and TCP API, and how it differs from the BSD Sockets API.
• How TLDK scales DOS prevention mechanisms such as SynCookie and SynCache.
• How TLDK scales TCP Timers to handle millions of connections per core.
• How TLDK integrates with Virtual Switches likes VPP?
• How will existing TCP and UDP based applications such as NGINX or Apache Traffic Server benefit from TLDK?

The Transport Layer Development Kit (TLDK) is new Layer 4 implementation, developed for DPDK (dpdk.org) and VPP (https://wiki.fd.io/view/VPP).

TCP and UDP socket based applications are core technologies in the modern data-center. The introduction of Orchestration, SDN controllers, vSwitches and Virtual Dataplane have seen a huge change in how network packets are delivered to data-centre nodes. While the stacks that consume these packets, the socket based applications have evolved much more slowly. Virtual switches like VPP and OVS-DPDK, have benefited from the vector packet processing technology in DPDK to scale process millions of packets per second, while TCP and UDP socket based applications have not yet benefited. There have been a number recent efforts to port various L4 implementations to DPDK, and while these usually give a modest performance benefit the performance of these is not yet revolutionary.

TLDK has been developed from the ground up to be a very fast and scalable UDP and TCP implementation, targeted for use with DPDK and VPP. In this presentation will provide an overview of TLDK design and implementation. This deep-dive will include:-

• Overview of the TLDK UDP and TCP API, and how it differs from the BSD Sockets API.
• How TLDK scales DOS prevention mechanisms such as SynCookie and SynCache.
• How TLDK scales TCP Timers to handle millions of connections per core.
• How TLDK integrates with Virtual Switches likes VPP?
• How will existing TCP and UDP based applications such as NGINX or Apache Traffic Server benefit from TLDK?