bat 0.3.4

Bro Analysis Tools

Bro Analysis Tools

The BAT Python package supports the processing and analysis of Bro data with Pandas, scikit-learn, and Spark

Why BAT?

Bro already has a flexible, powerful scripting language why should I use BAT?

Offloading: Running complex tasks like statistics, state machines, machine learning, etc.. should be offloaded from Bro so that Bro can focus on the efficient processing of high volume network traffic.

Data Analysis: We have a large set of support classes that help bridge from raw Bro data to packages like Pandas, scikit-learn, and Spark. We also have example notebooks that show step-by-step how to get from here to there.

Example: Pull in Bro Logs as Python Dictionaries

from bat import bro_log_reader
...
    # Run the bro reader on a given log file
    reader = bro_log_reader.BroLogReader('dhcp.log')
    for row in reader.readrows():
        pprint(row)

Output: Each row is a nice Python Dictionary with timestamps and types properly converted.

{'assigned_ip': '192.168.84.10',
'id.orig_h': '192.168.84.10',
'id.orig_p': 68,
'id.resp_h': '192.168.84.1',
'id.resp_p': 67,
'lease_time': datetime.timedelta(49710, 23000),
'mac': '00:20:18:eb:ca:54',
'trans_id': 495764278,
'ts': datetime.datetime(2012, 7, 20, 3, 14, 12, 219654),
'uid': 'CJsdG95nCNF1RXuN5'}
...

Example: Bro log to Pandas DataFrame (in one line of code)

from bat.log_to_dataframe import LogToDataFrame
...
    # Create a Pandas dataframe from a Bro log
    bro_df = LogToDataFrame('/path/to/dns.log')

    # Print out the head of the dataframe
    print(bro_df.head())

Output: All the Bro log data is in a Pandas DataFrame with proper types and timestamp as the index

                                                     query      id.orig_h  id.orig_p id.resp_h \
ts
2013-09-15 17:44:27.631940                     guyspy.com  192.168.33.10       1030   4.2.2.3
2013-09-15 17:44:27.696869                 www.guyspy.com  192.168.33.10       1030   4.2.2.3
2013-09-15 17:44:28.060639   devrubn8mli40.cloudfront.net  192.168.33.10       1030   4.2.2.3
2013-09-15 17:44:28.141795  d31qbv1cthcecs.cloudfront.net  192.168.33.10       1030   4.2.2.3
2013-09-15 17:44:28.422704                crl.entrust.net  192.168.33.10       1030   4.2.2.3

More Examples

• Easy ingestion of any Bro Log into Python (dynamic tailing and log rotations are handled)

• Bro Logs to Pandas Dataframes and Scikit-Learn

• Dynamically monitor files.log and make VirusTotal Queries

• Dynamically monitor http.log and show ‘uncommon’ User Agents

• Running Yara Signatures on Extracted Files

• Checking x509 Certificates

• Anomaly Detection

• See BAT Examples for more details.

Analysis Notebooks

BAT enables the processing, analysis, and machine learning of realtime data coming from Bro.

• Bro to Scikit-Learn: Bro to Scikit

• Bro to Matplotlib: Bro to Plot

• Bro to Parquet to Spark: Bro->Parquet->Spark

• Bro to Kafka to Spark: Bro->Kafka->Spark

• Clustering: Picking K (or not): Clustering K Hyperparameter

• Anomaly Detection Exploration: Anomaly Detection

• Risky Domains Stats and Deployment: Risky Domains

Install

$ pip install bat

Documentation

bat-tools.readthedocs.org

Thanks

• The DummyEncoder is based on Tom Augspurger’s great PyData Chicago 2016 Talk