Graph-go — zero config, full visibility

What it is

A new open-source tool called graph-info promises an instant, interactive map of your stack: databases, tables, services, storage buckets — all in one live graph. It has been reported that the tool auto-discovers infrastructure by talking to the Docker daemon and watching Kubernetes informers, and allegedly no manual inventory is required. The UI shows topology, dependency links and real-time health updates via WebSockets — every five seconds, so you can see changes as they happen.

How it works

Graph-info probes common components: PostgreSQL (tables and FKs), MySQL, MongoDB, Redis keyspaces, Elasticsearch shards, S3/MinIO buckets and plain HTTP services. It classifies running Docker containers, extracts credentials where possible, and watches Docker events for topology changes; Kubernetes discovery covers Namespaces, Deployments, StatefulSets, DaemonSets and Pods. Think of it as a quick way to stop guessing which service talks to which database. Relief, finally.

Try it (and a quick caveat)

Want to give it a spin? The project ships as backend and frontend containers on GitHub and recommends mounting the Docker socket for auto-discovery (docker run -v /var/run/docker.sock:/var/run/docker.sock …). Pre-built binaries and a Docker Compose demo with sample Postgres, Mongo and MinIO are provided for fast testing. One important note: mounting the Docker socket works — but it’s a known security trade-off, so treat that access with care in production.

Why it matters

Observability is the hot coin of the realm right now, and tools that reduce manual inventory and surface implicit dependencies are useful — especially for teams wrestling with hybrid stacks and fast-moving containers. Whether graph-info becomes a staple or a handy stopgap depends on adoption and how it handles edge cases, but for a quick visual of “what talks to what,” it’s worth a look.

Sources: github.com/guilherme-grimm, Hacker News