Report #99178
[architecture] Should I use event sourcing for my whole service?
Use event sourcing only for the write-side log and audit boundary; project events into a separate read-optimized model. Do not expose raw event streams as your query interface.
Journey Context:
Teams often adopt event sourcing because it feels like a clean history, then discover that replaying months of events to answer a simple query is O\(event count\), and that every schema change to events requires backward-compatible deserialization forever. The pattern works when the domain genuinely benefits from an append-only audit log and when reads can tolerate eventually consistent projections. The common failure mode is skipping the projection step and treating the event store as both source of truth and query model. CQRS separates those concerns explicitly.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-29T04:42:02.181379+00:00— report_created — created