Power Metrics Flask Server

This project provides a simple Flask server that returns CPU power metrics from pyJoules. The server is containerized using Docker.

Project Structure

```

. ├── code │ ├── flaskserver.py │ └── powerscraper.py ├── docker │ └── Dockerfile └── README.md

````

• code/flask_server.py : The main Flask application that defines a /api/cpu_power endpoint.
• code/power_scraper.py : A helper module using pyJoules to measure CPU energy usage.
• docker/Dockerfile : Dockerfile to build the container.

Prerequisites

• Docker installed on your system.

Build the Docker Image

1. Navigate to the root directory of this project (the directory containing docker/ and code/).
2. Run the Docker build command, specifying the Dockerfile location:

bash docker build -t power-metrics -f docker/Dockerfile .

This will:

- Pull the Python 3.12 slim base image,
- Copy the `code/` directory,
- Install necessary Python dependencies (`pyJoules`, `requests`, `flask`),
- Expose port 5000,
- Specify the startup command (`CMD`).

Run the Docker Container

After the image is built, you can run a container:

bash docker run -d --privileged -p 5000:5000 --network=host power-metrics

• -d runs the container in the background (detached).
• -p 5000:5000 maps host port 5000 to container port 5000.
• --name power-metrics-container sets a friendly container name.

If you want to override the default port, you can set the PORT environment variable:

bash docker run -d --privileged -p 8080:8080 -e PORT=8080 --network=host power-metrics

Now, the Flask server will listen on port 8080 inside the container, and the host machine will map 8080 to the container port.

Testing the Endpoint

You can test the /api/cpu_power endpoint using curl or any REST client:

```bash curl -X POST http://localhost:5000/api/cpu_power

or, if you used a custom port (say 8080):

curl -X POST http://localhost:8080/api/cpu_power ```

You should receive a JSON response with power metrics like:

json { "timestamp": 1672839007.12345, "tag": "waste_time_sleep", "duration": 0.1, "package_0": <energy value>, "dram_0": <energy value>, ... }

Environment Variables

• DEVICE_ID: If needed for your environment, set this when running the container using -e DEVICE_ID=<id>; the power_scraper.py uses this to build commands for measuring power on a specific device (default is "0").
• PORT: Specify which port Flask should listen on (default is 5000).

Notes & Tips

1. Device ID: Since your power_scraper class constructor requires a device_id, you should either provide it explicitly or let it default to '0'. If you truly need a dynamic device ID, you can pass it as an environment variable at runtime via -e DEVICE_ID=<id>.
2. Debugging: If you encounter any issues with pyJoules, ensure your hardware and OS are supported. You may also need privileges or special environment settings for hardware-level measurement.
3. Logs: You can view the container logs via docker logs power-metrics-container for troubleshooting.

With these changes and instructions, you should have a fully working Flask server Docker image that can be built and run successfully.

Enjoy measuring power metrics with your Flask application!