SWIMming in the microservices ocean

on YouTube ↗28m

" summary

Microservices become cooperative by combining a gossip membership (SWIM) with consistent hashing for application‑level sharding, not by more HTTP. Concrete stack: RingPop for routing and sharding, Hyperbahn for service discovery and circuit breaking, TChannel for RPC and Zipkin for tracing. Uber reports handling 20 million events per second. The claim: this combo keeps membership convergent, routes requests deterministically, and scales with predictable rebalancing.

AI-written summary · about

" chapters (4)

1. 0:00Start
2. 10:58SCALABLE WEAKLY CONSISTENT INFECTION STYLE MEMBERSHIP PROTOCOL
3. 15:29WHY? WHAT?
4. 18:26APPLICATION-LEVEL SHARDING

" related talks

• 34mSWIMming in the Microservices OceanLuca Maraschi · Node Week 2017
• 35mBuilding Scalable Stateful ServicesCaitie McCaffrey · Strange Loop 2015
• 59mMicroservice Swarms: Decentralized Discovery and ScalingAllen Holub · NDC Oslo 2017
• 27mSolving service discovery: Node.js microservices without a registryRichard Rodger · Microservices Day London 2016
• 52mBeyond Microservices: Streams, State and ScalabilityGwen Shapira · DevTernity 2022
• 45mResponsible MicroservicesNathaniel Schutta · Voxxed Days Microservices 2019
• 27mSimplifying Building and Managing Microservices with MicroAsim Aslam · Golang UK 2016
• 26mBuilding Scalable Microservice APIs From The BeginningJyrone Parker · APIdays San Francisco 2019