Host Welcome

Vikram Venkataraghavan

Open Remarks

Jim Xu TSC Chair, LFEdge General Board Member

Oleg Berzin TSC Co-Chair

Keynotes from LFEdge

Tina Tsou Board Chair, LF Edge 

K8s Edge storage with Lightbits and Sherlock

Abstract: In this talk, we will use Sherlock, an open source performance framework to run and stress databases on K8s (MySQL, PostgreSQL, SQLserver and MongoDB). The Lightbits CSI driver is utilizing NVMe/TCP as the transport. NVMe over TCP was added to the Linux kernel in version 4.10 and has been maintained by a growing community of developers from various companies. For the storage part we will use Lightbits software to create the fastest possible software defined storage solution for K8s.

Sagy Volkov, Distinguished Performance Architect, Lightbits

Daniil Egranov System Architect, Arm

BIO: Daniil Egranov is a System Architect in the Arm Architecture Technology Group. He is part of the Infrastructure Security Team and working on platform and firmware security architectures. His main focus is researching security solutions for cloud and edge platforms.  

Imran Yusuf, Director Hardware Ecosystem, Arm

Cilium introduction and improvement for DSR

Jiang Wang, Linux Kernel Engineer, Bytedance

Abstract: 

Cilium is a eBPF based open source software for Kubernetes and Docker containers. By using eBPF, Cilium can achieve a very high scale with low overhead. In this talk, we will first introduce eBPF and then Cilium networking functions. We will also discuss SNAT and Direct Server Return (DSR) modes and how to improve DSR.

BIO: Jiang Wang is a Linux Kernel Engineer of System Technologies and Engineering team at ByteDance. He has been working on Linux Kernel for more than 10 years. He works in various areas for upstreaming Linux, including security, virtualization and networking. His recent work involves eBPF and Cilium.

ORAN

Abstract: 

By disaggregating RAN deployments, 5G presents great opportunities as well challenges to build disaggregated multi-vendor solutions. ORAN Alliance is standardizing the orchestration and management of RAN networks so that vendors can comply with these standards to build standardized multi-vendor solutions. 

In this talk I will give a brief overview of activities in ORAN working groups and somewhat details about the Service Management Orchestration (SMO) and ORAN Cloud (O-Cloud)

Sukhdev Kapur Senior Distinguished Engineer, Juniper

BIO: Sukhdev Kapur is Distinguished Engineer at Juniper Networks. He is part of CTO organization and is driving the Contrail architecture, 5G, Edge, and Cloud Native initiatives. Sukhdev is a networking veteran with over 20 years experience in highly available distributed systems, cloud computing, virtualization, disaster recovery, policy based mobile workloads management, and software defined networks. He has held architectural positions in Arista Networks, F5, Cisco, Alteon, and others. He holds several patents in cloud computing, hierarchical data center deployments, cloud based disaster recovery, high availability, data center fabric automation,  etc.

Vehicle Computing: Vision and Challenges

Abstract: Vehicles have been majorly used for transportation in the last century. With the proliferation of onboard computing and communication capabilities, we envision that future connected vehicle (CVs) will be serving as a mobile computing platform in addition to their conventional transportation role for the next century. In this article, we present the vision of Vehicle Computing, i.e., CVs are the perfect computation platforms, and connected devices/things with limited computation capacities can rely on surrounding CVs to perform complex computational tasks. We also discuss Vehicle Computing from several aspects, including several key and enabling technologies, case studies, open challenges, and the potential business model. 

Weisong Shi, Professor, IEEE Fellow

Robotics Edge HPC - Human safety and emergency commands are utmost priority for robot operation. To achieve reliable voice command response in mobile / SWaP-constrained robots – in the presence of servo motor noise, background conversations and sounds, and intermittent Internet dropout – requires high performance computing under severe size and power constraints. Noise removal, false positive resilience, variability in human commands, and personnel / proprietary sound privacy requirements pose difficult challenge, while at the same time utilizing cloud connectivity for containerization, orchestration, and maintenance telemetry. For the CPS Robot SSES Blueprint, Signalogic is adding 20,000 word vocabulary onboard ASR (automatic speech recognition) to demonstrate the onboard HPC approach

Jeff Brower, Signalogic

Introduction of Nephio

Kandan Kathirvel Product Lead - Telco Cloud & Orchestration, Google

BIO: An established industry leader in delivering results in Product development, technology & telecommunications solutions, across strategy, and research & innovation with 20+ years of Telecommunications, Cloud & Datacenter experience. Former Akraino TSC Chair.

Host Welcome

Practice and Thinking of CFN (Computing Force Network) in China Mobile

Abstract: Introduces the new concept, new goal and vision of China Mobile's Computing Force Network (CFN), as well as including China Mobile's research progress and practice in multi-cloud management, ubiquitous computing force scheduling etc.

hanyu ding

PCIe Net enable Object Storage at Edge

As the edge computing prevailing, the number of edge clusters are growing in million scales. The traditional mega data center with millions servers located in one site evolve to millions of small clusters worldwide. Hence we need a new solution not only account for this architecture shift, but also for energy constrains.

In this talk, we will introduce a novel PCIe-based system-level interconnect model and arm based micro-server cluster to make clusters more compact and energy-saving. As the practical application, we use object storage as an example and show how well the energy efficiency this architecture can achieve.

Keywords: PCIe Net, object storage, micro-server cluster, edge computing.

Bio: Dr. Fu Li (LEO) Leo Li
the pioneer of cloud native application on PCIe-based networking system. Dr. Fu Li comes with experience on cloud computing, HPC, named-data networking and system-level architecting and algorithm optimization.

China Unicom's 5G MEC and Private Network Practices

Abstract: China Unicom's practices on MEC and 5G Private Network will be introduced.

Rong Huang

OpenStack at the Edge

Ildiko Vancsa Senior Manager, Open Infrastructure Foundation

BIO: As a senior manager I have a strong strategic mindset with an attention to detail and over 10 years of experience in technical leadership. An experienced speaker, I have been on stage in front of thousands of people, work with media and analysts, create strategic content for blogs and articles and can address both executives and developers at their level.

Highlights of LF Edge/Akraino Cloud Game White Paper

Davy Zhang

Software Director, Y-semi Computing

Break (Lunch time at APAC)

Akraino Release 7 Planning

Jim Xu Oleg Berzin

Akraino Subcommittee Annual Report

Jeff Brower

PCEI (PCEI Architecture Document)

Public Cloud Service Providers and 3rd-Party MEC Providers are deploying Edge instances to better serve their end-users and applications, A multitude of these applications require close inter-working with the MNO Edge deployments to provide predictable latency & throughput, reliability, and other telco-grade requirements. The purpose of this blueprint family is to specify a standard set of APIs to expose towards Public Cloud and 3rd-Party MEC Service Provider instances at the Edge.

Oleg Berzin Fellow, Equinix

BIO: Technology leader with extensive expertise in creating, defining, developing and executing complex multi-technology programs focusing on forward-looking vision, innovation and revenue-driving results. Over 20 years of experience in Wireless and Wireline Communications, Internetworking and Product Development. Unique combination of leadership, strategic thinking, extensive practical work and individual expert technical skills.

Close Remarks