Speaker Deck

Reliable Crash Detection and Failover with Orchestrator

by Shlomi Noach

Published June 3, 2016 in Programming

The nature of MySQL replication implies various crash scenarios of varying availability impact.
Orchestrator is an open source project that discovers, manages and recovers your MySQL replication.
This talk discusses how Orchestrator detects failures with minimal false positives/negatives, and figures out the best method of recovery even in complex topologies.

- Complex topology types and crash scenarios
- Common crash detection methods
- Configuration based vs. State based recoveries
- The complexity of promotion paths
- Potential post-recovery limbo states
- Flapping & acknowledgements
- Visibility & control