AMPLI-SYNC is an open-source synchronization framework designed for bidirectional data replication between a local SQLite database (edge/client) and a centralized relational database (PostgreSQL, MS SQL, MySQL, Oracle).

The framework enables offline-first architectures where applications operate independently of network availability and synchronize automatically once connectivity is restored.

https://github.com/AMPLIFIER-sp-z-o-o/ampli-sync-rn-example

https://github.com/AMPLIFIER-sp-z-o-o/custom-tomcat

Clients operate on a fully functional local SQLite database.

Characteristics:

• No runtime dependency on network
• Full transactional consistency locally
• Local writes never blocked by remote availability
• Deterministic sync recovery after failure

The synchronization engine operates as a separate logical layer:

Application Layer
        ↓
Local Database (SQLite)
        ↓
Sync Adapter
        ↓
Transport Layer (HTTPS / Queue / RPC)
        ↓
Server Sync Endpoint
        ↓
Central Database

The engine does not enforce schema structure. It operates on metadata and change tracking strategies defined per project.

To enable synchronization, each table must be initialized, look for:

public void InitSync(String schema)

AMPLI-SYNC uses incremental synchronization based on:

• Version counters
• Logical delete markers

Two synchronization directions:

Client sends:

• New records
• Updated records
• Soft-deleted records

Server validates: - Schema integrity - Referential integrity - Business rules

Server returns:

• Records modified since client's last sync watermark
• Conflict resolutions (if any)
• Forced overrides (optional)

AMPLI-SYNC does not enforce a single strategy.

Supported models:

Version-based resolution.

Server overrides client in conflict.

Edge device takes priority.

Domain-specific resolution layer (recommended for enterprise systems).

Each sync batch operates inside transactional boundaries:

• Atomic batch processing
• Partial retry support

Server endpoints must be idempotent to support retry mechanisms.

• Horizontal scaling behind load balancer
• Stateless sync endpoints
• Batch processing

• Batched payload transmission
• Delta-based updates
• Local indexing for change queries

• Compression (gzip)
• JSON streaming or binary encoding
• Parallel table sync
• Watermark-based incremental queries

Recommended security measures:

• HTTPS only
• Token-based authentication (JWT / API keys)
• Request signature validation
• Payload size limits
• Rate limiting
• Audit logging

AMPLI-SYNC supports:

• Exponential backoff retry
• Resume from last successful watermark
• Idempotent reprocessing
• Network interruption tolerance

Failure scenarios handled:

• Partial batch failure
• Network timeout
• Duplicate submission
• Concurrent modification

For SaaS systems:

• Tenant isolation at DB level (schema-based or row-level security)
• Tenant-aware sync tokens
• Scoped synchronization contexts

AMPLI-SYNC architecture has been validated in:

• Mobile field sales systems
• Field service applications
• Distributed B2B ordering platforms
• Multi-tenant SaaS ecosystems

The framework is optimized for high write-frequency environments with thousands of records per sync cycle.

AMPLI-SYNC is not a product. It is a synchronization framework requiring:

• Domain modeling
• Sync strategy design
• Conflict definition
• Infrastructure preparation

It is intended for engineering teams building scalable distributed systems.

MIT License

If you plan to implement AMPLI-SYNC in a commercial or enterprise environment, architectural and integration support is available.