DeployGuard CLI (deploy-rollback-cli)

A lightweight, fault-tolerant Command-Line Interface (CLI) built with modern C++ for automated software deployment and intelligent state rollback.

This tool is designed to act as a reliable state-machine for CI/CD pipelines, ensuring that any deployment failure triggers an immediate and clean rollback to a safe state, minimizing system downtime.

✨ Key Features

• Automated Deployment Execution: Seamlessly triggers and monitors external deployment scripts or binaries.
• Intelligent Rollback Mechanism: Real-time capturing of process exit codes. Automatically halts and initiates recovery scripts if a non-zero exit status is detected.
• Process Sub-system Monitoring: Built to directly interface with the OS environment using standard C/C++ libraries.
• Extensible Scripting Foundation: Decouples core logic from execution logic via flexible shell scripting (deploy.sh, rollback.sh).

📁 Repository Structure

.
├── CMakeLists.txt        # CMake build configuration
├── .github/workflows/    # CI (build + unit tests)
├── include/
│   └── deploy-cli.h      # Core headers and definitions
├── scripts/
│   ├── build.sh          # Automated compilation script
│   ├── deploy.sh         # Mock deployment execution script
│   ├── rollback.sh       # Mock system recovery script
│   └── setup_v1.sh       # Demo environment bootstrap
├── src/
│   └── main.cpp          # CLI entry point and state-machine logic
└── tests/
    └── test_main.cpp     # doctest unit tests

🚀 Getting Started

Prerequisites

• A C++17 compatible compiler (GCC, Clang, or MSVC)
• CMake (3.10 or higher)

Build Instructions

You can quickly build the project using the provided shell script:

chmod +x scripts/build.sh
./scripts/build.sh

Alternatively, use standard CMake commands:

mkdir build && cd build
cmake ..
make

Usage

Run the CLI tool by specifying the target deployment path:

./deploy-cli --path /var/www/target_app

Expected Workflow (Simulated):

1. The CLI parses the arguments and invokes scripts/deploy.sh.
2. If deploy.sh encounters an error (e.g., exit 1), the CLI intercepts the error.
3. The CLI immediately triggers scripts/rollback.sh to restore the system state.
4. Logs are outputted to the standard output/error stream for pipeline monitoring.

Exit Codes

deploy-cli distinguishes three outcomes so CI/CD pipelines can react differently to each:

Code	Meaning
0	Deployment succeeded. No rollback was necessary.
1	Deployment failed, but the automatic rollback succeeded — the system is back to a stable, known-good state.
2	Rollback itself failed (or a crash-recovery rollback failed) — the system is in an unsafe/unknown state and needs manual intervention.

Running Tests

Unit tests (doctest) cover ConfigParser::parseYaml, ArgumentParser::parseArgs, and ProcessManager/DeployController — including shell-injection safety (malicious target_path values), the fork/exec hang-avoidance behavior, and all three exit-code outcomes above:

./scripts/build.sh
cd build && ctest --output-on-failure

🛠️ Future Roadmap

• Replace std::system with POSIX popen for real-time stdout/stderr streaming.
• Implement a persistent logging system to write deployment histories to .log files.
• Add JSON configuration parsing for complex deployment workflows.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.