Product · Sync
AthenaDB Sync
AthenaDB Sync is the planned two-way replication model for enrolled Lite and Server replicas. Each replica can continue using its local database offline, then exchange and reconcile changes after reconnecting. Existing object-level oplog and changeset foundations remain in progress; scoped replication and AthenaDB Convergent Record Versioning are Planned / Early Access designs, not GA capabilities.
Offline work, then two-way reconciliation
Connectivity is not required for continued local use. An enrolled replica records changes locally while offline; when connectivity returns, AthenaDB Sync is designed to reconcile eligible changes in both directions.
Whole or explicitly scoped
The approved design supports whole-database replication and explicitly scoped replication. Scoped replication is Planned / Early Access and is not an implemented or GA claim.
AthenaDB Convergent Record Versioning
A planned reconciliation model that merges non-overlapping field changes, chooses a deterministic winner when both replicas change the same field, and preserves recoverable version history instead of silently discarding the losing value.
- Changes to different fields merge into one converged record.
- Concurrent changes to the same field use a deterministic winner on every replica.
- Recoverable version history retains prior values for inspection or restoration.
Identity before replication
A sync peer is an enrolled replica with its own cryptographic identity, not an anonymous endpoint. Enrollment, scope, and revocation ride the Instance Trust Fabric so replication is explicit and policy-governed.

