Data Flow
The data flow in Eidetica follows a structured sequence of interactions between core components.
Basic Flow
- User creates a Instance with a database backend
- User creates Databases within the database
- Operations construct immutable Entry objects through EntryBuilder
- Entries reference parent entries, forming a directed acyclic graph
- Entries are stored and retrieved through the database interface
- Authentication validates and signs entries when configured
Authentication Flow
When authentication is enabled, additional steps occur during commit:
- Entry signing with cryptographic signatures
- Permission validation for the operation type
- Bootstrap handling for initial admin configuration
- Verification status assignment based on validation results
This ensures data integrity and access control while maintaining compatibility with unsigned entries.
CRDT Caching Flow
The system uses an efficient caching layer for CRDT state computation:
- Cache lookup using Entry ID and Store as the key
- On cache miss, recursive LCA algorithm computes state and caches the result
- Cache hits return instantly for subsequent queries
- Performance scales well due to immutable entries and high cache hit rates