Distributed Load Tester & API Benchmarker

A high-performance, containerized load-testing suite written in Go and orchestrated with Kubernetes. This system benchmarks API reliability by distributing concurrent load generation across worker nodes and aggregating real-time telemetry via gRPC into a stateful persistence layer.

Core Capabilities

• Massive Concurrency: Leverages Go’s Goroutines and channels to execute thousands of simultaneous HTTP requests with a non-blocking fan-in pattern.
• High-Fidelity Observability: Calculates exact P95 and P99 latency percentiles to provide deep insights into system tail latency.
• Low-Latency Telemetry: Utilizes a custom gRPC networking contract for type-safe, high-speed batch data transmission between workers and the aggregator.
• Stateful Storage: Persists every benchmark run and aggregated metric into PostgreSQL for historical reporting.

Architecture

The system follows a distributed producer-consumer model orchestrated by Kubernetes:

1. Worker Nodes (Go): Deployed as Kubernetes Jobs. They read dynamic environment variables, execute concurrent HTTP load, and stream batched results.
2. Central Aggregator (Go): A Kubernetes Deployment (gRPC server) that consumes telemetry, calculates percentiles, and manages database transactions.
3. Persistence Layer (PostgreSQL): A stateful service ensuring all benchmark data survives container lifecycle events.

Tech Stack

• Language: Go (Golang)
• Protocols: gRPC, Protocol Buffers (Protobuf)
• Containerization: Docker (Multi-stage Alpine builds)
• Orchestration: Kubernetes (Deployments, Services, Jobs)
• Automation: PowerShell (Configuration Templating)

Quick Start (Local Development)

Execute the following block to initialize infrastructure, build the worker engine, and run a dynamic benchmark:

# 1. Initialize Infrastructure (Database & Aggregator)
kubectl apply -f postgres.yaml
kubectl apply -f aggregator.yaml

# 2. Build the Worker Engine Image
docker build -t worker-service -f services/worker/Dockerfile.worker .

# 3. Execute Dynamic Benchmark Run via Automation Script
# The script templates worker.yaml with your URL and Request Count
# Usage: .\stress.ps1 -url "<TARGET_URL>" -count <NUMBER_OF_REQUESTS>
powershell.exe -File .\stress.ps1 -url "http://www.google.com" -count 100

# 4. Analysis (View P95/P99 metrics and DB insertions)
kubectl logs deployment/aggregator-deployment