Stateful App: Efficient Dispatching
summary
The talk examines when stateless architectures become inefficient for Uklon’s services and identifies specific workloads that make stateful designs preferable. It covers techniques for scaling stateful services—partitioning and replication—addresses disaster recovery concerns (RTO and RPO), and describes a migration path from actor-based solutions to native stateful implementations.
AI-written summary · about
chapters (10)
- 0:00Stateful apps intro
- 3:10Reliability in real-time
- 6:47Driver modes and filters
- 9:56Key workloads
- 12:53Stateful app challenges
- 17:34KV storage pros/cons
- 23:40Scaling strategies
- 28:20Consistency across replicas
- 32:20State recovery and optimization
- 35:36Next steps
related talks
38mExploring Stateful Microservices in the Cloud Native World
29mStateful Workloads in Kubernetes: A Deep Dive
44mDesign Patterns for Distributed Systems in #AKS
54mExploring Stateful Microservices in the Cloud Native World
34mDistributed applications and Kubernetes
35mStateful Apps On Kubernetes - Bring Them On!
35mHands-On Stateful Serverless Applications with K8s and Stateful Functions
20mApplication Modernization: From Concept to Kubernetes
