I built a small tool to answer a question I’ve asked too many times: is this production ready?

#devops #showdev #sre #tooling

I’ve been in a lot of release conversations where everything technically looked fine.

The app deployed.
The tests passed.
Dashboards existed.
Alerts existed.
Logs existed.

But still, someone would ask:

“Are we actually production ready?”

And the answer was usually some version of:

“I think so.”

That always bothered me.

Because production readiness is not just whether something can deploy. It is whether the system can survive the boring, painful things that happen after it goes live:

a Kafka broker goes down
consumer lag starts growing
a Kubernetes pod keeps restarting
Terraform creates something too open
storage grows faster than expected
retries amplify a downstream database issue
nobody knows which risk should be fixed first
So I started building Beacon.

Beacon is a local production-readiness scanner for distributed systems.

The idea is simple: give Beacon your infra/config/runtime inputs, and it tries to answer:

Is this safe to release?
What are the biggest operational risks?
What should be fixed first?
What evidence is missing?
Right now it supports things like Kafka configs, Kubernetes YAML, Terraform/static infra, Schema Registry-style checks, ACL checks, runtime snapshots, and flow scenarios like:

API -> Kafka -> Consumer -> Database
It has a local UI and CLI, and runs with Docker:

docker run --rm -p 8765:8765 ghcr.io/mishraricha1806/beacon:latest ui --host 0.0.0.0 --port 8765
Then open:

http://127.0.0.1:8765/
I’m intentionally sharing it early because I don’t want to build this in a vacuum.

If you work in DevOps, SRE, platform engineering, backend systems, Kafka, Kubernetes, or release engineering, I’d genuinely love your feedback.

Especially:

What checks would you expect from a production-readiness tool?
What risks do teams usually miss before release?
What would make this useful in a real CI/CD or release process?
What should Beacon support next?
The principle I’m trying to keep is:

Deterministic checks first. Clear operational reasoning. No black-box magic.

It’s still early, but I’d love for people to try it and tell me what’s missing.