Azure Container Apps Performance Testing Harness

This repository provides an opinionated, end-to-end harness for performance testing web/API workloads on Azure Container Apps. It bundles infrastructure-as-code templates, a Python-based CLI orchestrator, Azure Load Testing integrations, and a simulation app so customers can iterate on cost and performance quickly.

Choose Your Path

Method	What You Need	Time to First Test
AZD Deployment	Azure CLI + AZD	~15 minutes
Docker Quickstart	Docker Desktop	~5 minutes
AI-Guided Setup	Copilot Chat + Azure CLI	~10 minutes

Prerequisites

• Python 3.11+ (for the CLI harness)
• Azure CLI with an active subscription (az login)
• Docker Desktop (for the local quickstart only)
• AZD CLI (optional — for azd up deployment)

Quick Start

Option 1: AZD (deploy to Azure in one command)

# Install AZD if you haven't: https://aka.ms/azd
azd auth login
azd up
# Choose an environment name and region when prompted

That's it — AZD provisions the Container Apps environment, simulation app, and Azure Load Testing resource. The post-deploy hook prints your app URL and next steps.

Option 2: CLI harness (full control)

# Clone and set up
git clone https://github.com/simonjj/aca-scaler-lab.git
cd aca-scaler-lab
python -m venv .venv
source .venv/bin/activate      # Windows: .venv\Scripts\activate
pip install -r requirements.txt

# Generate a test config
cd src
python -m cli init --output my-test.yaml

# Edit my-test.yaml to set your container image and Azure settings, then:
python -m cli --config my-test.yaml run

Option 3: Docker (local, no Azure)

cd quickstart
docker compose up --build
# Dashboard at http://localhost:9090 after the test completes

Note: The default container image (simon.azurecr.io/simulation-app:latest) is the built-in simulation app. For custom workloads, push your image to your own ACR and update the config.

Project Layout

src/                # CLI, orchestrator, models, services
infrastructure/     # Bicep templates for Azure resources
infra/              # AZD-compatible infrastructure
load-tests/         # JMeter scenarios and ALT parameter files
simulation-app/     # Configurable FastAPI simulation target
scale-lab-ui/       # Web UI for guided multi-fabric tests
quickstart/         # Docker-based local preview (no Azure needed)
docs/               # Guides and reference material
tests/              # Unit and integration tests

Troubleshooting

Issue	Fix
az login fails	Run az account clear && az login
Deployment timeout	Check region availability for your workload profile
WorkloadProfile not found	Ensure all needed profiles are in the same deployment (multi-fabric bug fixed in Bicep)
Docker out of memory	Allocate ≥ 4 GB RAM to Docker Desktop

See docs/user-manual.md for the full CLI reference and configuration schema.