Iocane Cluster Monitor UI

This repository is a monitoring system for a cluster of raspberry pi's. A python script uses ssh to login to each raspberry pi in the network and retrieves their CPU usage info. It then writes the retrieved information to a json file (node_info.json). The interface also allows users to submit parallel jobs to the cluster using the slurm job scheduler.

Getting Started

In order to install and start a new monitoring cluster of raspberry pi's, here are the steps:

On the head node (raspberry pi):

1. Setup apache web server: sudo apt-get update && sudo apt-get install apache2 -y
2. git clone https://github.com/rubensl/Iocane-PiClusterManager.git
3. sudo mv Iocane-PiClusterManager/* /var/www/html
4. Configure cluster_info.json to correspond to cluster info.
5. Configure node_info.json to correspond to each node in the cluster adding in their hostnames, usernames and password.
6. Install required dependencies: sudo python3 setup.py build && sudo python3 setup.py install
7. Start up backend of monitoring application: sudo python3 start_server.py

Testing Web UI:

1. Go to preferred web browser and search http://[Head_Node_IP]