Scaling Rails API to Write-Heavy Traffic

on YouTube ↗56m

" summary

Mobile-game backends serve tens of millions of users and millions of DB transactions per minute by using multiple databases with sharding, manual read/write switching, DNS-based service discovery and weighted sampling for dynamic shard allocation. Rails apps run on MySQL with replication (Rails 6 read‑your‑write 2s default), resource-focused API responses include affected resources to avoid reads, and partial resharding or pool/push/double-write patterns keep locality. Practical additions: connection pooling fixes, at‑least‑once message queues, real multi‑DB load tests; result is high write scalability with tolerable eventual consistency.

AI-written summary · about

" chapters (11)

1. 0:00Start
2. 3:17Mobile Game API Backends: System Overview
3. 25:47Scenario: Join a Guild and Defeat a Raid Boss
4. 26:25Solution: Partial Resharding
5. 29:08Overview of 2 Phase Commit for Comparative Study
6. 32:02Reduce Risk of Cross-Shard Transaction Anomalies
7. 35:36Keep Inconsistency Observable and Acceptable
8. 38:12Keep Idempotency and Reach Eventual Consistency
9. 40:18Utilize At Least Once Message Queue for Strict Completion
10. 50:34Conduct Load Test before Service Launch
11. 53:53Summary: Scaling Rails API to Write Heavy Traffic

" related talks

• 29mScaling Rails to 2M MySQL requests per secondEugene Kenny · San Francisco Ruby Conference 2025
• 39mPostgres & Rails 6 Multi-DB: Pitfalls, Patterns, PerformanceGabe Enslein · RailsConf 2019
• 46mScaling With FriendsGeoffrey Dagley · Big Ruby 2013
• 31mA Rails performance guidebook: from 0 to 1B requests/dayCristian Planas · RubyConf TH 2022
• 14mDatabase Performance at Scale for RoR ApplicationsRocio Delgado · GORUCO 2016
• 20mSafely Decomposing a Highly Available Rails AppAdam Forsyth · Rocky Mountain Ruby 2015
• 45mTechnically, a TalkEileen M Uchitelle · rubyday 2020
• 29mA Rails Performance Guidebook: from 0 to 1B requests/dayCristian Planas · RailsConf 2022