AppDynamics Smart Agent Management with GitHub Actions

Automated deployment and lifecycle management of AppDynamics Smart Agent across multiple EC2 hosts using GitHub Actions with a self-hosted runner.

🎯 Overview

This lab demonstrates how to use GitHub Actions to manage AppDynamics Smart Agent and various AppDynamics agents (Node, Machine, DB, Java) across multiple Ubuntu EC2 instances within a single AWS VPC. The project includes 11 automated workflows covering the complete agent lifecycle.

Key Features:

• 🚀 Parallel Deployment - Deploy to multiple hosts simultaneously
• 🔄 Complete Lifecycle Management - Install, uninstall, stop, and clean agents
• 🏗️ Infrastructure as Code - All workflows version-controlled
• 🔐 Secure - SSH key-based authentication, private VPC networking
• 📈 Massively Scalable - Deploy to thousands of hosts with automatic batching
• 🎛️ Self-hosted Runner - Executes within your AWS VPC

📊 Architecture

All infrastructure runs in a single AWS VPC with a shared security group. The self-hosted runner communicates with target hosts via private IPs.

View detailed architecture diagrams →

🚀 Quick Start

Prerequisites

• AWS VPC with EC2 instances (Ubuntu)
• Self-hosted GitHub Actions runner in the same VPC
• SSH key pair for authentication
• AppDynamics Smart Agent package and config

1️⃣ Clone and Configure

git clone https://github.com/chambear2809/github-actions-lab.git
cd github-actions-lab

2️⃣ Set Up GitHub Secrets

Navigate to: Settings → Secrets and variables → Actions

Required Secret:

• SSH_PRIVATE_KEY - Your SSH private key (PEM format)

3️⃣ Set Up GitHub Variables

Navigate to: Settings → Secrets and variables → Actions → Variables

Required Variable:

• DEPLOYMENT_HOSTS - Target host IPs (one per line)

  172.31.1.243
  172.31.1.48
  172.31.1.5
  

Optional Variables:

• SMARTAGENT_USER - User for Smart Agent service (e.g., appdynamics)
• SMARTAGENT_GROUP - Group for Smart Agent service (e.g., appdynamics)

4️⃣ Deploy!

Via GitHub UI:

1. Go to Actions tab
2. Select "Deploy Smart Agent"
3. Click "Run workflow"
4. Optionally adjust batch size (default: 256)
5. Click "Run workflow"

Via GitHub CLI:

gh workflow run "Deploy Smart Agent" --repo chambear2809/github-actions-lab

# With custom batch size
gh workflow run "Deploy Smart Agent" --repo chambear2809/github-actions-lab -f batch_size=128

📋 Available Workflows

Deployment (1 workflow)

Workflow	Description	Scale	Trigger
Deploy Smart Agent (Batched)	Installs Smart Agent and starts service	Any	Manual only

Agent Installation (4 batched workflows)

Workflow	Command	Scale	Trigger
Install Node Agent (Batched)	smartagentctl install node	Any	Manual only
Install Machine Agent (Batched)	smartagentctl install machine	Any	Manual only
Install DB Agent (Batched)	smartagentctl install db	Any	Manual only
Install Java Agent (Batched)	smartagentctl install java	Any	Manual only

Agent Uninstallation (4 batched workflows)

Workflow	Command	Scale	Trigger
Uninstall Node Agent (Batched)	smartagentctl uninstall node	Any	Manual only
Uninstall Machine Agent (Batched)	smartagentctl uninstall machine	Any	Manual only
Uninstall DB Agent (Batched)	smartagentctl uninstall db	Any	Manual only
Uninstall Java Agent (Batched)	smartagentctl uninstall java	Any	Manual only

Smart Agent Management (2 batched workflows)

Workflow	Description	Scale	Trigger
Stop and Clean Smart Agent (Batched)	Stops service and purges data	Any	Manual only
Cleanup All Agents (Batched)	Deletes /opt/appdynamics directory	Any	Manual only

Total: 11 workflows - All batched workflows support configurable batch sizes (default: 256)

📚 Documentation

• Deployment Guide - Complete setup and configuration instructions
• Architecture Diagrams - Visual infrastructure and workflow diagrams

🛠️ How It Works

1. Developer pushes code or manually triggers a workflow
2. GitHub Actions receives the event and assigns job to self-hosted runner
3. Runner loads target hosts from GitHub variables
4. Parallel Execution - Runner SSHs into each target host simultaneously
5. Commands Execute - Install/uninstall/stop/clean operations run on each host
6. Results Reported - Success/failure status sent back to GitHub

🔐 Security

• Private Network - All communication via VPC private IPs
• SSH Keys - Stored securely as GitHub secrets
• No Public Access - Target hosts don't need public IPs
• Security Group - Restricts SSH access to runner only

📈 Scaling

All workflows use automatic batching to support any number of hosts:

How It Works

• Automatic batching - Splits hosts into groups of 256 (configurable)
• Sequential batch processing - Avoids overwhelming runner resources
• Parallel within batch - Each batch processes all hosts simultaneously
• Works at any scale - 1 host to thousands

Batching Strategy

1. Splits your host list into manageable batches
2. Processes each batch as a separate job sequentially
3. Deploys to all hosts within each batch in parallel using background processes

Examples:

• 10 hosts = 1 batch, all deploy in parallel
• 500 hosts = 2 batches × 256 hosts
• 1,000 hosts = 4 batches × 256 hosts
• 5,000 hosts = 20 batches × 256 hosts

🤝 Contributing

This is a lab/demo project. Feel free to fork and adapt for your own use cases!

🔗 Links

• GitHub Actions Documentation
• AppDynamics Documentation
• Self-hosted Runners

---

Built with ❤️ for AppDynamics automation