# 📡 Go Observability Collector

A lightweight and pluggable **Go library** for sending structured **logs** and **metrics** to your own gRPC endpoint. Ideal for embedding into any Go microservice or infrastructure component.

---

## ✨ Features

✅ Unified library for Logs and Metrics  
✅ Buffered and batch-based transmission  
✅ Automatic flush on interval or size  
✅ Graceful shutdown with `SIGINT` / `SIGTERM`  
✅ Rich metadata (hostname, service name, OS, caller info, etc.)  
✅ Pluggable into any Go project

---

## 📦 Installation

```bash
go get github.com/yourusername/observability-collector

Replace with your real module path or structure if you keep logs and metrics as separate packages like logger and metriccollector.

---

🔌 gRPC Backend Requirements

You’ll need a gRPC backend with two services:

✅ LogReceiver Service

service LogReceiver {
  rpc ReceiveLogs (LogBatch) returns (LogResponse);
}

✅ MetricReceiver Service

service MetricReceiver {
  rpc ReceiveMetrics (MetricBatch) returns (MetricResponse);
}

---

🛠️ Quickstart

1. Initialize Log Collector

import "your_module_path/logger"

logCollector := logger.NewLogCollector("localhost:50051", "my-service", 10, 5*time.Second)

logCollector.Info("Application started")
logCollector.Error("Something failed: %v", err)

2. Initialize Metric Collector

import "your_module_path/metriccollector"

metricCollector := metriccollector.NewMetricCollector("localhost:50052", "my-service", 10, 5*time.Second)

metricCollector.Collect("cpu.usage", metriccollector.GAUGE, 0.83, []string{"core:0"}, "percent")

---

📋 Metric Types

Type	Description
GAUGE	Value that can go up or down
COUNT	Cumulative counter
HISTOGRAM	Distribution of values
RATE	Rate of change over time
SUMMARY	Percentiles
METER	Events per time unit
DISTRIBUTION	Custom distributions
SET	Unique values over time

---

🧼 Graceful Shutdown

Both collectors handle termination (SIGINT, SIGTERM) and flush remaining entries:

🛑 Shutdown detected! Flushing remaining logs/metrics...
🚀 LogCollector stopped.
🚀 MetricCollector stopped.

Or call manually:

logCollector.Stop()
metricCollector.Stop()

---

📄 Example (Combined)

package main

import (
	"time"
	"your_module_path/logger"
	"your_module_path/metriccollector"
)

func main() {
	logs := logger.NewLogCollector("localhost:50051", "combined-service", 10, 5*time.Second)
	metrics := metriccollector.NewMetricCollector("localhost:50052", "combined-service", 10, 5*time.Second)

	logs.Info("🚀 Starting service")
	metrics.Collect("requests.total", metriccollector.COUNT, 1, []string{"route:/health"}, "req")

	time.Sleep(10 * time.Second)

	logs.Stop()
	metrics.Stop()
}

---

---

📃 License

MIT © madhavan-21

---

👥 Contributors

Built by madhavan-21, inspired by modern observability platforms.