The networking industry has witnessed the foundational shift of packet processing from proprietary hardware and software to open architecture including COTS (Commercial Off-The-Shelf) hardware and open-source software based solutions across hyperscale datacenter to enterprise edge and wireless networks. In order to meet the insatiable throughput requirement in networking datapath, there has been a strong push for open-source software based network applications and acceleration libraries running on general purpose CPUs. It is represented by DPDK, FD.io/VPP, Arm RAL, Snort/Vectorscan, etc. Arm is deeply involved in these foundational projects to ensure support and optimization for the latest features of the Arm architecture.
Since the introduction of Arm Neoverse, the deployment of networking functions on Arm has accelerated significantly and are now available in servers, DPUs, security appliances and wireless infrastructure. More developers are looking to deploy their networking applications and services onto Arm based platforms to benefit from the leading performance and power efficiency nature of the Arm architecture. Common challenges networking developers usually face are how to quickly build their networking applications on top of existing well-optimized open-source components on Arm, as well as having a quantitative performance target they can expect for a specific set of use cases.
Dataplane Stack is an integrated software stack that pulls together different open-source components, providing a reference solution for building networking application and user cases with fine-tuned configuration on Arm Neoverse platform. The Dataplane Stack project includes a well-documented user guide and set of scripts to accelerate the deployment on various Arm-based systems.
Objectives:
Use case list:
Support platform list:
Dataplane Stack is continuously updated and integration tested following a regular release cadence. It is made available to build from source. So far we have delivered three releases (latest release: 24.03) of Dataplane Stack, incorporating new features and adding new supported platforms along the release cycle.
Figure 1: Dataplane stack roadmap
If you are interested in using or contributing to Dataplane Stack, you can find all the resources you need on our GitLab repository. You will have access to the source code, user guide documentation, scripts and more. You can also report issues, request features, and submit pull requests.
Join us to ask questions, share feedback, and connect with other users and developers of the Dataplane Stack.
Get Started with Dataplane Stack