ELECT: Enabling Erasure Coding Tiering for LSM-tree-based Storage
Offered By: USENIX via YouTube
Course Description
Overview
Explore a 22-minute conference talk from USENIX FAST '24 that introduces ELECT, a novel distributed key-value store designed to optimize storage efficiency and performance in tiered architectures. Learn how ELECT leverages erasure coding and replication techniques in conjunction with log-structured merge trees to address the challenges of skewed workloads in practical key-value storage systems. Discover how this innovative approach achieves significant storage savings in the hot tier while maintaining high performance and data availability guarantees compared to traditional replication-based systems like Cassandra. Gain insights into ELECT's hotness-aware data conversion and offloading mechanisms, as well as its tunable parameter for balancing storage savings and access performance.
Syllabus
FAST '24 - ELECT: Enabling Erasure Coding Tiering for LSM-tree-based Storage
Taught by
USENIX
Related Courses
MongoDB for DBAsMongoDB University Optimizing Performance for SQL Based Applications
Microsoft via edX App Deployment, Debugging, and Performance
Google Cloud via Coursera Application Deployment, Debug, Performance 日本語版
Google Cloud via Coursera Optimize TensorFlow Models For Deployment with TensorRT
Coursera Project Network via Coursera