Event Sourcing - What Could Possibly Go Wrong?

Drawing on work across three commercial products, this talk examines practical challenges in implementing event sourcing. It covers event storage, schema evolution, scalability, eventual consistency, and common failures such as broken read models and duplicate events.

Syllabus

Intro
Software Mill
What is Event Sourcing?
Why Event Sourcing?
ES and CQRS level 3.1.1
ES and CQRS alternatives Change Capture Data CDC logging instead of event bus?
library vs. framework
Event store for Akka Persistence
Cassandra perfect for ES? partitioning by design
Event serialization
Event payload
State replay time
Memory consumption
Fixing state
Handling duplicates
Broken read model
Multi aggregate transactional update
Optimistic compensation action
Cluster = split brain
Cluster best practises
Summary