报 告 人：曹通 院长，上海焜耀研究院

报告时间：09月26日（周一）13:00～17:00

报告地点：腾讯会议（485-745-522）

邀 请 人：余航 讲师

报告摘要：

To avoid single point of failure and enhance system’s dependability, some mission-critical computer systems (e.g., Google’s Chubby lock service, Boeing’s ARINC 659 SAFEbus) deploy multiple core components (replicas) in the network and implement Byzantine fault tolerant protocols. In case of f arbitrary faults, such BFT protocols guarantee that the system can persistently execute correct orders via a number of replicas (from f+1 to 3f+1 depending on different settings). In general, classic BFT protocols are designed to tolerate faults in a small-scale network, for instance, Google’s Chubby lock service only consists of 5 nodes. In the past decade, many efforts have been made to explore the possibilities of BFT protocols in a large-scale computing network for different purposes, such as, Hyperledger fabric, Facebook’s DiemBFT. Despite some improvements, it remains an open challenge to tolerate Byzantine faults in a large-scale computing network, in particular, the network that requests high throughput and low latency. In this talk, several BFT protocols are introduced along with some applications.

报告人简介:

曹通，上海焜耀研究院，科研负责人。卢森堡大学博士，博士论文被提名为卢森堡大学年度杰出博士论文。主要研究方向为分布式系统容错性，计算机网络共识机制，网络安全。2022年协议实验室（Protocol Labs）RFP-X 科研基金获得者。长期研究基于IPFS传输协议的Filecoin区块链分布式存储网络的技术服务平台，集群架构与算法相结合，具有丰富的实战经验。