Change data capture is quietly becoming the backbone of modern data platforms, and Debezium with Kafka Connect is the open-source stack most teams reach for. It is powerful — but opaque. When a connector stalls during its initial snapshot, or a task dies silently at 3 a.m., teams usually find out from downstream reports that are suddenly stale, not from the pipeline itself.
So we built a tool for exactly that:
- a Java wrapper around Debezium / Kafka Connect that watches each connector through the two phases that matter — initialization (snapshot and first streaming, where most pipelines get stuck) and steady-state operation;
- it tracks connector health continuously and exposes it as JMX metrics, so the pipeline state lands in the monitoring you already run — our dashboards live in Grafana;
- and it is designed to grow into more than a monitor: the roadmap is a tool that can diagnose CDC pipeline failures and, eventually, auto-fix them — driven by an AI module.
The idea is simple: a pipeline that moves your data should be able to tell you how it is doing — before your reports do.
Running or planning CDC pipelines (Debezium, Kafka Connect, PostgreSQL logical replication)? See the Data Integration service — or just talk to us.