How to write a custom script to monitor a mining rig?

Understanding the Purpose of a Mining Rig Monitoring Script

A custom script to monitor a mining rig serves multiple purposes, including tracking hardware performance, detecting anomalies, and ensuring optimal operation. Whether you're managing one or multiple rigs, having real-time data is essential for maintaining efficiency. The core idea behind writing such a script is to automate repetitive tasks like checking GPU temperatures, hash rates, and system uptime. By using programming languages like Python or Bash, you can extract this data from your mining software and operating system tools.

Monitoring scripts help identify issues before they become critical.

Selecting the Right Tools and Libraries

Before diving into code, ensure you have the appropriate tools installed on your mining machine. For example, if you're using Linux, you'll likely need nvidia-smi for NVIDIA GPU monitoring or amdgpu-top for AMD GPUs. Additionally, Python libraries such as psutil for system-level information and GPUtil for GPU details are highly useful. On Windows, tools like Task Scheduler, PowerShell, and WMI (Windows Management Instrumentation) provide access to hardware metrics.

• Install nvidia-smi or amdgpu-top depending on your GPU brand
• Use pip to install GPUtil and psutil in your Python environment
• Ensure mining software like NiceHash or Claymore exposes an API for data retrieval

Accessing Mining Software APIs

Most modern mining applications expose internal statistics via local HTTP APIs. For instance, NiceHash Miner offers a RESTful API that allows querying miner status, current algorithm, and device health. To use it, you typically send a GET request to http://localhost:port/api/status. The response is usually in JSON format, making it easy to parse with Python or any scripting language.

APIs allow real-time access to mining performance without manual checks.

To integrate with your script:

• Find the correct port used by your miner’s API
• Send a request using Python's requests library
• Parses relevant fields like hashrate, temperature, and fan speed

Writing the Core Monitoring Logic

Once the data sources are identified, the next step is to structure your script. Begin by importing necessary modules like json, requests, time, and os. Then create functions to fetch GPU stats and miner stats separately. Use try-except blocks to handle potential errors like unreachable APIs or missing binaries.

Here's a basic outline of what your Python script might include:

import jsonimport requestsimport GPUtilimport psutilimport time





def get_gpu_stats():





gpus = GPUtil.getGPUs()
for gpu in gpus:
    print(f'GPU ID: {gpu.id}, Load: {gpu.load * 100}%, Temp: {gpu.temperature}°C')
def get_miner_stats(api_url):
 
try:
    response = requests.get(api_url)
    data = response.json()
    # Extract and display relevant miner info
except Exception as e:
    print('Miner API error:', e)
while True:
 
get_gpu_stats()
get_miner_stats('http://localhost:38080/api/status')
time.sleep(60)  # Run every minute
The loop ensures continuous monitoring at set intervals.
 
Logging and Alerting Mechanisms
Beyond just displaying data, a robust monitoring script should log historical data and trigger alerts when thresholds are breached. You can append results to a CSV file for later analysis. Implement alerting by sending notifications through email, Discord, or Telegram using webhooks.
 
For logging:
 
Open a file in append mode
Write timestamps along with relevant metrics
Flush after each write to prevent data loss
For alerts:
 
Set temperature or hashrate thresholds
Use smtplib for email alerts
Use webhook URLs for instant messaging platforms
Alert systems help prevent overheating and downtime automatically.
 
Frequently Asked Questions
Can I run the monitoring script on a remote server instead of the mining rig?

Yes, but you’ll need SSH access or exposed APIs from the mining rig. Remote execution may introduce latency and security considerations.

How do I customize the frequency of monitoring?Adjust the sleep duration inside the while loop. For example, changing time.sleep(60) will control how often the script runs.

What permissions are needed to access GPU data?On Linux, you may need sudo privileges to read certain GPU metrics, especially with tools like nvidia-smi. On Windows, running the script as administrator might be required.

Is there a lightweight alternative to Python for this task?Yes, shell scripts using commands like nvidia-smi --query-gpu=index,name,temperature.gpu,utilization.gpu --format=csv can offer similar functionality with lower resource usage.