Episode

Supervised State Replication in Elixir with Micah Cooper

Podcast: Elixir Wizards
Published: May 7, 2026
Duration seconds: 2820
Processing state: processed
Canonical source: https://smartlogic.fireside.fm/s15-e02-supervised-state-replication-elixir-micah-cooper
Audio: https://aphid.fireside.fm/d/1437767933/03a50f66-dc5e-4da4-ab6e-31895b6d4c9e/ed9443b4-16a8-4cf2-ac69-af6e6b9bfe4a.mp3
JSON: /v1/public/podcasts/elixir-wizards/episodes/supervised-state-replication-in-elixir-with-micah-cooper
Markdown: /podcast/elixir-wizards/supervised-state-replication-in-elixir-with-micah-cooper.md

Actions

POST https://stenobird.com/v1/public/podcasts/elixir-wizards/episodes/supervised-state-replication-in-elixir-with-micah-cooper/transcription-requests
Idempotently request low-priority transcript generation for this episode.
GET https://stenobird.com/podcast/elixir-wizards/supervised-state-replication-in-elixir-with-micah-cooper.md
Read the agent-friendly Markdown representation of this episode resource.

Summary

Explore the mechanics of state replication in Elixir using the Viewstamps algorithm. Micah Cooper discusses building Visor, a library designed to replicate GenServer state across distributed nodes for enhanced fault tolerance.

Topics

Elixir
Distributed Systems
Data Replication
GenServer
Viewstamps Algorithm
Event Sourcing
Fault Tolerance
Visor Library
BEAM

Highlights

Main idea: Visor uses the Viewstamps replication algorithm to synchronize GenServer state across multiple nodes
Practical takeaway: Distinguishing between distributed systems and data replication is critical for designing reliable Elixir applications
Failure mode: Event sourcing can introduce significant complexity and memory overhead if history management and snapshotting are not handled
Technical insight: Replicating a Write-Ahead Log (WAL) is a foundational pattern for achieving consistency in distributed databases
Design philosophy: Building for curiosity and experimentation allows developers to push the boundaries of the BEAM ecosystem

Chapters

1:00 Developer Background: Micah discusses his transition from Ruby to Elixir and his early experiences in the ecosystem.
4:35 Inspiration from TigerBeetle: How the design principles of the TigerBeetle database influenced the development of Visor.
8:10 Replicating GenServer State: A deep dive into the mechanics of updating and persisting state within a GenServer context.
11:55 Scaling Replication: The benefits and trade-offs of increasing the number of backup nodes in a replication algorithm.
15:20 Orchestration and Redundancy: Using orchestration platforms to manage synchronized and redundant nodes.
22:20 Distributed Systems vs. Data Replication: Clarifying the fundamental differences between these two often-confused concepts.
25:45 Challenges in Event Sourcing: Discussing the industry resistance to event sourcing due to its inherent complexity.