tldr: Posted this anomaly detector that uses ChatGPT at https://github.com/krypted/Lightweight-GPT-Log-Anomaly-Detector. Can be used untrained or with lightly trained troves of macOS system logs. Tweak to your delight… Use A number of options may need to be run in a given environment. To do so, use the following: Create a virtual environment python3.6 -m virtualenv venv Activate the virtual environment source venv/bin/activate Install the requirements pip install -r requirements.txt Create the log file sudo log show --last 10m > log-info-10m.txt Run the script log-analysis.py to parse log file and generate anomalous logs. Example usage: python log-analysis.py --log_filename log-info-10m.txt python log-analysis.py --log_filename log-info-10m.txt --use_error_keywords True --score_threshold 0.5 python log-analysis.py --log_filename log-info-10m.txt --use_error_keywords…
-
-
Mac OS X Log Analysis
Mac OS X has a lot of options for logging data. But these options can actually cause a lot of white noise. So there are two tools that you should look at if you’re interested. The first is Splunk: http://www.splunk.com/LogAnalysis The second is Sawmill: http://www.sawmill.net