DEPART - Replica Decoupling for Distributed Key-Value Storage
Offered By: USENIX via YouTube
Course Description
Overview
Explore a 16-minute conference talk from FAST '22 that introduces DEPART, a novel approach to distributed key-value storage. Learn about the concept of replica decoupling, which separates the storage management of primary and redundant copies of replicas in distributed key-value stores. Discover how this technique mitigates I/O costs in indexing and provides tunable performance. Examine the innovative two-layer log design that enables adjustable ordering for redundant copies, resulting in balanced read/write performance. Gain insights into the implementation of DEPART as a prototype built on Cassandra and its performance improvements over traditional systems. Understand how DEPART achieves significant throughput enhancements for writes, reads, scans, and updates under various consistency levels and parameter settings.
Syllabus
FAST '22 - DEPART: Replica Decoupling for Distributed Key-Value Storage
Taught by
USENIX
Related Courses
Advanced Operating SystemsGeorgia Institute of Technology via Udacity High Performance Computing
Georgia Institute of Technology via Udacity GT - Refresher - Advanced OS
Georgia Institute of Technology via Udacity Distributed Machine Learning with Apache Spark
University of California, Berkeley via edX CS125x: Advanced Distributed Machine Learning with Apache Spark
University of California, Berkeley via edX