Why is Edge-to-Cloud Data Sync so Hard?
Data generated at the source (IoT gateways, edge servers) must be persisted locally due to shaky networks, but eventually needs to reach central platforms for analysis.
Generic edge databases lack time-series optimizations, and setting up separate ETL pipelines (like Kafka Connect) on constrained edge hardware is complex, resource-intensive, and brittle.
Ingest Once. Query Locally. Sync Reliably.
MonoTS unifies local persistence with outbound data synchronization. Written in Rust for a minimal resource footprint, it provides an optimized SQL engine for local analytics and a built-in CDC service that guarantees your data reaches its destination — even through network outages.
Core Pillars
Optimized for Edge Hardware
Single-process binary. Configurable memory budgets. Efficient Parquet storage on local media. Speaks SQL over gRPC. Built for ARM and x86 gateways.
Native Data Synchronization (CDC)
Treat tables as ordered logs. SQL-managed CREATE STREAM continuously pushes changes to Kafka and Data Lakes without separate ETL jobs on the gateway.
Resilience for Shaky WAN
Keep writing locally during outages. CDC streams automatically checkpoint progress and resume with at-least-once delivery when the link returns.
One Engine, Multiple Destinations
MonoTS natively synchronizes data to:
Message Queues
Apache Kafka — real-time row events in JSON for downstream stream processing.
Data Lakes
Delta Lake (Parquet + _delta_log) and local filesystem — sealed batches for historical analytics.