We're excited to share Librarian, an open-source database replicator built for the cloud-native era.
The Problem We're Solving:
Traditional CDC solutions like Kafka Connect require JVM runtime, complex connector management, and expose generic infrastructure metrics that
don't help you debug data pipelines. We built Librarian to change that.
What Makes Librarian Different:
Single Binary - No JVM, no external dependencies, no connector management. Just run it.
Truly Lightweight - Runs on modest hardware with minimal resource overhead
Pipeline-First Observability - Built-in stats server that shows what matters: events processed, bytes transferred, and error counts—not
garbage collection metrics
Native Replication - Uses MongoDB Change Streams and PostgreSQL logical replication for efficient, real-time CDC
Debezium Compatible - Drop-in replacement for existing Debezium consumers
Supported Today:
- Sources: MongoDB, PostgreSQL
- Targets: Kafka, S3 (Parquet), Local filesystem
Get Started in Minutes:
Stream MongoDB changes to Kafka with a single command. No configuration files, no cluster setup—just a simple URL-based connection string and
you're streaming data.
Librarian is open source (MIT license). We'd love your feedback as we continue building.
Check it out: github.com/turbolytics/librarian
No comments yet.