watonomous.github.io

[ Electrical Division : Server Monitoring ]

Created by [ Rowan Dempster] on Dec 27, 2019

Overview

The main purpose of monitoring the vitals of the rugged is to determine how often the rugged is being used as a server and if it is throttled by lack of flash memory. Furthermore, analysis can be done on this data to highlight particularly productive hours or track which processes are the most intensive.

Background

https://pypi.org/project/psutil/

Implementation

All code related to server monitoring can be found here

The server_monitoring.py script can be run to log details about the temperature, CPU usage, and RAM usage of the server. It updates a csv file with this information every 5 minutes, and can be configured to have a longer or shorter period. This script should be automatically run upon booting up the rugged, and the local csv file it writes to should be pushed to a remote server at the end of every day.

The temperature_test.py script can be run as a benchmark for the rugged to test how well the liquid cooling performs. The script loads up all 88 cores on the rugged to 100% usage and periodically writes the temperature of the two core sectors to a text file. If the temperature approaches critical level the script will automatically terminate.

Future Work

A script needs to be created to periodically push this data to a server and clear the local data.

Document generated by Confluence on Nov 28, 2021 22:40

Atlassian