One VTOrc to Rule Them All - High Availability in a Distributed Database System

Explore high availability in distributed database systems through this conference talk on VTOrc, the next-generation cluster management service for Vitess. Dive into the architecture, design principles, and implementation of VTOrc, a feature that ensures scalability and durability in Vitess, a distributed database system built around MySQL. Learn about unplanned leader election, durability policies, semi-sync replication, and cross-cell durability. Discover how VTOrc handles various failure scenarios while respecting user-specified durability rules. Watch a demonstration of VTOrc's capabilities in multiple failover scenarios and gain insights into its successful production deployments, including at PlanetScale. Understand the theory behind VTOrc's implementation and its significance in achieving high availability in distributed database environments.

Intro
Architecture
Problem Statement
Design Principles
Unplanned Leader Election
Durability Policies & Semi-Sync
Semi-Sync Durability
Cross-Cell Durability
Custom Durability Policies
More Failure Scenarios
Resources