The Raft Consensus Algorithm

Diego Ongaro and John Ousterhout

October 2017

Source code available at https://github.com/ongardie/raft-talk.
Unless otherwise noted, this work is:
© 2012-2017 Diego Ongaro, © 2012-2014 John Ousterhout.
Licensed under the Creative Commons Attribution 4.0 International License.

How Is Consensus Used?

Top-level system configuration

Replicate entire database state

Paxos Protocol

• Leslie Lamport, 1989
• Nearly synonymous with consensus

“The dirty little secret of the NSDI community is that at most five people really, truly understand every part of Paxos ;-).”
—NSDI reviewer

“There are significant gaps between the description of the Paxos algorithm and the needs of a real-world system...the final system will be based on an unproven protocol.”
—Chubby authors

Raft's Design for Understandability

We wanted an algorithm optimized for building real systems

• Must be correct, complete, and perform well
• Must also be understandable

“What would be easier to understand or explain?”

• Fundamentally different decomposition than Paxos
• Less complexity in state space
• Less mechanism

Raft User Study

Raft Implementations

LogCabin

• Started as research platform for Raft at Stanford
• Developed into production system at Scale Computing
• Network service running Raft replicated state machine
• Data model: hierarchical key-value store, kept in memory
• Written in C++ (gcc 4.4's C++0x)