This application is designed to collect and analyze logs from remote machines hosted on Amazon Web Services (AWS) and other cloud hosting services.
Note: This application was specifically designed for use with Pastel Network's log files. However, it can be easily adapted to work with any log files by modifying the parsing functions, data models, and specifying the location and names of the log files to be downloaded. It is compatible with log files stored in a standard format, where each entry is on a separate line and contains a timestamp, a log level, and a message. The application has been tested with log files several gigabytes in size from dozens of machines and can process all of it in minutes. It is designed for Ubuntu 22.04+, but can be adapted for other Linux distributions.
To adapt this application for your own use case, refer to the included sample log files and compare them to the parsing functions in the code. You can also modify the data models to store log entries as desired.
The application consists of various Python scripts that perform the following functions:
- Connect to Remote Machines: Using the boto3 library for AWS instances and an Ansible inventory file for non-AWS instances, the application establishes SSH connections to each remote machine.
- Download and Parse Log Files: Downloads specified log files from each remote machine and parses them. The parsed log entries are then queued for database insertion.
- Insert Log Entries into Database: Uses SQLAlchemy to insert the parsed log entries from the queue into an SQLite database.
- Process and Analyze Log Entries: Processes and analyzes log entries stored in the database, offering functions to find error entries and create views of aggregated data based on specified criteria.
- Generate Network Activity Data: Fetches and processes network activity data from each remote machine.
- Expose Database via Web App using Datasette: Once the database is generated, it can be shared over the web using Datasette.
The tool is compatible with both AWS-hosted instances and any list of Linux instances stored in a standard Ansible inventory file with the following structure:
all:
vars:
ansible_connection: ssh
ansible_user: ubuntu
ansible_ssh_private_key_file: /path/to/ssh/key/file.pem
hosts:
MyCoolMachine01:
ansible_host: 1.2.3.41
MyCoolMachine02:
ansible_host: 1.2.3.19
(Both can be used seamlessly.)
To simplify the code, the tool is designed to delete all downloaded log files and generated databases each time it runs. Consequently, this can consume significant bandwidth depending on your log files' size. However, the design's high level of parallel processing and concurrency allows it to run quickly, even when connecting to dozens of remote machines and downloading hundreds of log files.
Designed for Ubuntu 22.04+, first install the requirements:
python3 -m venv venv
source venv/bin/activate
python3 -m pip install --upgrade pip
python3 -m pip install wheel
pip install -r requirements.txt
You will also need to install Redis:
sudo apt install redis -y
And install Datasette to expose the results as a website:
sudo apt install pipx -y && pipx ensurepath && pipx install datasette
To run the application every 30 minutes as a cron job, execute:
crontab -e
And add the following line:
*/30 * * * * . $HOME/.profile; /home/ubuntu/automatic_log_collector_and_analyzer/venv/bin/python /home/ubuntu/automatic_log_collector_and_analyzer/automatic_log_collector_and_analyzer.py >> /home/ubuntu/automatic_log_collector_and_analyzer/log_$(date +\%Y-\%m-\%dT\%H_\%M_\%S).log 2>&1