Data Access and Data Sharing in Microservices
summary
Five concrete ways to access data owned by other services: REST calls (simple but adds latency and tight coupling), data replication (copy columns into your DB — fast joins but consistency headaches), in-memory replicated caches using Hazelcast/Apache Ignite/GemFire/Coherence/Infinispan (low latency but memory and cold‑start limits), data sidecars/distributed Redis caches (delegate reads, reduce interruptions, still external dependency), and shared data / broader bounded contexts or monolith fallback (fast but makes changes harder and increases coupling).
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:03Data hardest part of microservices
- 3:28Database per service challenges
- 8:08Introducing data domains
- 14:31Accessing data you don’t own
- 18:27Direct RESTful calls
- 18:49Data replication
- 21:16In-memory replicated cache
- 26:13Data sidecar with cache
- 32:34Sharing data directly
- 35:11Choosing the right option
- 36:02Q&A
related talks
53mData Access and Data Sharing in Microservices
51mData management in a Microservices world
44mA Software Engineer's Journey Through Microservices Data Sharing
43mAutonomous microservices don't share data
47mArchitectural Patterns for Caching Microservices
27mData Consistency in Microservices Architecture
51mOne Approach to Implementing Data Discovery in a Microservice Architecture
30mDatabases in the Microservices World