Prometheus Proxy HTTP

This project implements a custom HTTP proxy in Python that receives metric data in Prometheus remote_write format and forwards it to the real server configured via environment variables.

---

🎯 Why a custom proxy?

While Prometheus can send metrics directly, this proxy allows:

• Injecting custom logic before forwarding
• Inspecting, logging, or transforming metrics on the fly
• Adding a security or access control layer if needed

---

🧠 Architecture

[ Prometheus Remote Write Client ]
              |
              v
   [ Prometheus HTTP Proxy (Python) ]
              |
              v
[ Real Prometheus in Kubernetes (monitoring namespace) ]

---

🚀 Running Locally

Requirements

• Python 3.9+
• pip install -r requirements.txt

Environment variables (.env)

The proxy uses the following environment variables for configuration:

Variable	Default Value	Description
REAL_PROM_REMOTE_WRITE	http://prometheus-operator-kube-p-prometheus.monitoring.svc.cluster.local:9090/api/v1/write	URL of the real Prometheus remote_write endpoint
PROXY_PORT	8080	Port where the proxy listens

---

🐳 Docker

docker build -t proxy-prometheus .
docker run -p 8080:8080 --env-file .env proxy-prometheus

---

☸️ Kubernetes Deployment

Kubernetes manifests are provided in the k8s/ directory to deploy the proxy as a service.

Relevant files

• k8s/deployment.yaml: Deployment configuration with necessary environment variables
• k8s/service.yaml: NodePort service exposing port 8080

Exposed Ports

Service	Protocol	Container Port	NodePort	Description
otel-prometheus-proxy	HTTP	8080	30080	Endpoint for Prometheus remote_write

---

🧪 Example Usage

The proxy listens for POST requests at:

POST http://<host>:8080/write
Content-Type: application/x-protobuf

Compatible with clients sending data in Prometheus remote_write format.