Power TiKV with In-Memory Engine - Enhancing Performance and Latency
CNCF [Cloud Native Computing Foundation] via YouTube
Overview
Explore the implementation of an in-memory engine to enhance TiKV's performance in this 36-minute conference talk by Chenjie Tang from PingCAP. Learn how this innovative approach addresses the challenges of unpredictable read latency in distributed key-value databases, particularly when dealing with multiple key-value versions resulting from numerous UPDATE operations in transactional mode. Discover how the in-memory engine not only stabilizes low latency for TiKV but also improves overall performance in scenarios with hot ranges and heavy read workloads. Gain insights into the technical aspects of this solution and its impact on TiKV's efficiency as a distributed database system.
Syllabus
Power TiKV with in-Memory Engine | 使用内存引擎强化 TiKV - Chenjie Tang, PingCAP
Taught by
CNCF [Cloud Native Computing Foundation]