Network security system

Patent metadata

Field	Value
Publication number	US-9699205-B2
Application number	US-201514841634-A
Country	US
Kind code	B2
Filing date	Aug 31, 2015
Priority date	Aug 31, 2015
Publication date	Jul 4, 2017
Grant date	Jul 4, 2017

Abstract

A security platform employs a variety techniques and mechanisms to detect security related anomalies and threats in a computer network environment. The security platform is “big data” driven and employs machine learning to perform security analytics. The security platform performs user/entity behavioral analytics (UEBA) to detect the security related anomalies and threats, regardless of whether such anomalies/threats were previously known. The security platform can include both real-time and batch paths/modes for detecting anomalies and threats. By visually presenting analytical results scored with risk ratings and supporting evidence, the security platform enables network security administrators to respond to a detected anomaly or threat, and to take action promptly.

First claim

What is claimed is:

1. A network security system comprising:
a computation engine implemented using Apache Storm or Apache Spark Streaming, configured to receive unbounded first event data indicative of activity on a computer network, to detect first indicia of possible security breaches in a real-time processing mode based on the first event data, and to generate real-time analysis result data representing the first indicia for output to a user;
an Apache Hadoop framework including a Hadoop Distributed File System (HDFS) to store the real-time analysis result data and second event data indicative of activity on the computer network, the second event data having been stored in the HDFS prior to analysis of the first event data by the computation engine; and
an Apache Spark cluster computing engine operatively coupled to the computation engine and the Apache Hadoop framework, and configured to operate concurrently with the computation engine, the Apache Spark cluster computing engine further configured to retrieve, from the HDFS, the real-time analysis result data and the second event data, and to detect, in a batch mode, second indicia of possible security breaches based on the second event data and the real-time analysis result data.

2. The network security system of claim 1 , wherein the first event data and the second event data each include machine data.

3. The network security system of claim 1 , wherein the first event data and the second event data each include timestamped machine data.

4. The network security system of claim 1 , wherein the computation engine is further configured to use outputs of the Apache Spark cluster computing engine, in conjunction with the first event data, to detect the first indicia of possible security breaches.

5. The network security system of claim 1 , wherein the first event data is a portion of an unbounded stream of event data.

6. The network security system of claim 1 , wherein the real-time path further includes a data intake and preparation engine configured to receive the first event data and the second event data from a plurality of heterogeneous data sources in the computer network, and to perform preprocessing of the first event data and the second event data before the first event data and the second event data are provided to the computation engine and the Apache Spark cluster computing engine, respectively; wherein the preprocessing includes at least one of: parsing the first and second event data, enriching the first and second data, and filtering the first and second event data.

7. The network security system of claim 1 , further comprising an Apache Kafka message broker to receive the first event data and the second event data and to pass the first event data to the computation engine and to pass the second event data to the Apache Spark cluster computing engine.

8. The network security system of claim 1 , wherein the real-time path further includes:
a data intake and preparation engine configured to receive the first event data and the second event data from a plurality of heterogeneous data sources in the computer network, and to perform preprocessing of the first event data and the second event data before the first event data and the second event data are provided to the computation engine and the Apache Spark cluster computing engine, respectively; and
an Apache Kafka message broker to receive the preprocessed first event data and second event data from the data intake and preparation engine and to pass the preprocessed first event data to the computation engine and to pass the preprocessed second event data to the Apache Spark cluster computing engine.

9. The network security system of claim 1 , wherein the computation engine includes a real-time anomaly detection engine and a real-time threat detection engine, wherein some of the first indicia of possible security breaches are detected by the real-time anomaly detection engine as security related anomalies and others of the first indicia are detected by the real-time threat detection engine as security related threats based on the detected anomalies.

10. The network security system of claim 1 , wherein the computation engine includes a real-time anomaly detection engine and a real-time threat detection engine, wherein some of the first indicia of possible security breaches are detected by the real-time anomaly detection engine as security related anomalies and others of the first indicia are detected by the real-time threat detection engine as security related threats based on the detected anomalies; and
wherein the Apache Spark cluster computing engine includes a batch anomaly detection engine and a batch threat detection engine, wherein some of the second indicia of possible security breaches are detected by the batch anomaly detection engine as security related anomalies and others of the second indicia are detected by the batch threat detection engine as security related threats based on the anomalies detected by the batch anomaly detection engine.

11. The network security system of claim 1 , wherein the second event data includes a larger amount of data than the first event data and has been generated over a longer time period than the first event data.

12. The network security system of claim 1 , wherein:
the computation engine executes a first plurality of versions of a plurality of machine learning models to detect the first indicia of possible security breaches in the real-time mode; and
the Apache Spark cluster computing engine executes a second plurality of versions of said plurality of machine learning models to detect the second indicia of possible security breaches in the batch mode.

13. The network security system of claim 1 , wherein:
the computation engine and the Apache Spark cluster computing engine collectively execute a plurality of machine learning models to detect the first indicia and second indicia of possible security breaches; and
the computation engine and the Apache Spark cluster computing engine share a model state of a particular machine learning model of the plurality of machine learning models.

14. The network security system of claim 1 , wherein:
the computation engine and the Apache Spark cluster computing engine collectively execute a plurality of machine learning models to detect the first indicia and second indicia of possible security breaches; and
a result from the Apache Spark cluster computing engine is used to update a model state of a machine learning model used by the computation engine.

15. The network security system of claim 1 , wherein:
the computation engine and the Apache Spark cluster computing engine collectively execute a plurality of machine learning models to detect the first indicia and second indicia of possible security breaches; and
a result from the computation engine is used to update a model state of a machine learning model used by the Apache Spark cluster computing engine.

16. The network security system of claim 1 , wherein:
the computation engine and the Apache Spark cluster computing engine collectively execute a plurality of machine learning models to detect the first indicia and second indicia of possible security breaches;
a result from the Apache Spark cluster computing engine is used to update a model state of a machine learning model used by the computation engine; and
a result from the computation engine is used to update a model state of a machine learning model used by the Apache Spark cluster computing engine.

17. The network security system of claim 1 , wherein the Apache Had

Assignees

• Splunk Inc

Inventors

• Muddu Sudhakar
• Tryfonas Christos
• Bulusu Ravi Prasad

Classifications

• G06N20/20 Primary

  Ensemble learning · CPC title

• H04L63/1416

  Event detection, e.g. attack signature detection · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F40/134

  Hyperlinking · CPC title

• G06N20/00

  Machine learning · CPC title