In 2021, UC Berkeley proposed the concept of Sky Computing . The so-called sky computing is a layer on the cloud platform, and its goal is to open up the isolated clouds and expand the distributed system from the data center to the global scope. Xline is an open source distributed kv storage system, implemented based on the curp consensus algorithm, designed to provide cross-cloud metadata management functions. Compared with traditional consensus protocols Raft or Paxos, curp has better performance in sky computing scenarios.
This Thursday (February 2nd) at 8:00 pm, Datan Technology and Distributed Labs cooperated, and Zhao Jiawei, Rust distributed storage engineer of Datan Technology, talked about the consensus algorithm and metadata management of cross-data center consistency, and introduced Xline How to achieve high-performance data consistency management across data centers.
Live highlights :
1. Introduction to Xline - the background and existing challenges of sky computing, the status of Xline.
2. Introduction to the curp consensus algorithm ——Compared with traditional consensus protocols, such as Raft and Paxos, how does the curp protocol achieve fast path.
3. The challenges encountered in the implementation of Xline - how to resolve the conflict between the revision mechanism and curp disorder semantics.
4. Xline's next optimization direction.
Please add the WeChat account of the distributed laboratory assistant in the poster to sign up to participate in the live broadcast.
The open source distributed KV storage Xline launched by DatenLord aims at multi-data center scenarios, realizes high-performance cross-cloud and cross-data center shared access to data, and ensures data consistency, which facilitates business systems to realize multi-center and multi-center Multi-active deployment.
GitHub link: Https: http://github.com/datenlord/Xline