Network security system with real-time and batch paths

• US 9,900,332 B2
• Filed: 06/07/2017
• Issued: 02/20/2018
• Est. Priority Date: 08/31/2015
• Status: Active Grant

• Current Assignee: Splunk Inc. (Cisco Systems, Inc.)
• Original Assignee: Splunk Inc. (Cisco Systems, Inc.)
• Inventors: Muddu, Sudhakar, Tryfonas, Christos, Bulusu, Ravi Prasad
• Primary Examiner(s): SHEHNI, GHAZAL B
• Application Number:
  US15/616,889
• Publication Number:
  US 20170272458A1
• Time in Patent Office: 258 Days
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/254 Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/444 Spatial browsing, e.g. 2D maps, 3D or virtual spaces
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 3/0482 Interaction with lists of selectable items, e.g. menus
• CPC Class Codes: G06F 3/0484 for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
• CPC Class Codes: G06F 3/04842 Selection of displayed objects or displayed text elements G06F3/0482 takes precedence
• CPC Class Codes: G06F 3/04847 Interaction techniques to control parameter settings, e.g. interaction with sliders or dials
• CPC Class Codes: G06F 40/134 Hyperlinking
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/20 Ensemble learning
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06V 10/225 based on a marking or identifier characterising the area
• CPC Class Codes: H04L 2463/121 Timestamp
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/145 involving simulating, designing, planning or modelling of a network
• CPC Class Codes: H04L 41/22 comprising specially adapted graphical user interfaces [GUI]
• CPC Class Codes: H04L 43/00 : Arrangements for monitoring or testing data switching networks
• CPC Class Codes: H04L 43/045 : for graphical visualisation of monitoring data
• CPC Class Codes: H04L 43/062 : related to network traffic
• CPC Class Codes: H04L 43/08 : Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
• CPC Class Codes: H04L 63/06 : for supporting key management in a packet data network cryptographic mechanisms or cryptographic arrangements for key management H04L9/08
• CPC Class Codes: H04L 63/1408 : by monitoring network traffic monitoring network traffic per se H04L43/00
• CPC Class Codes: H04L 63/1416 : Event detection, e.g. attack signature detection
• CPC Class Codes: H04L 63/1425 : Traffic logging, e.g. anomaly detection
• CPC Class Codes: H04L 63/1433 : Vulnerability analysis
• CPC Class Codes: H04L 63/1441 : Countermeasures against malicious traffic countermeasures against attacks on cryptographic mechanisms H04L9/002
• CPC Class Codes: H04L 63/20 : for managing network security; network security policies in general filtering policies H04L63/0227

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.

Claims

1. A network security system comprising:

a computation engine implemented using Apache Storm or Apache Spark Streaming, configured to receive 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;

a non-volatile storage system to store the real-time analysis result data and second event data indicative of activity on the computer network; and

an Apache Spark cluster computing engine operatively coupled to the computation engine and the non-volatile storage system, the Apache Spark cluster computing engine further configured to retrieve, from the non-volatile storage system, 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 second event data has been stored in the non-volatile storage system prior to analysis of the first event data by the computation engine.

3. The network security system of claim 1, wherein the non-volatile storage system comprises 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.

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

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

6. 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.

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

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;

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.

9. 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.

10. 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.

11. 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.

12. 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.

13. 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.

14. 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.

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

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.

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; 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.

17. 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.

18. 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.

19. The network security system of claim 1, further comprising:

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; and

a Redis cache to store the preprocessed first event data and second event data.

20. The network security system of claim 1, wherein:

the first event data and the second event data include timestamped machine data;

the computation engine is implemented on a first distributed processing platform and 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

the Apache Spark cluster computing engine is implemented on a second distributed processing platform and 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;

the computation engine and the Apache Spark cluster computing engine collectively execute a plurality of machine learning models to detect, respectively, the first indicia and the 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.

21. A method comprising:

detecting, in a real-time processing mode, first indicia of possible security breaches based on first event data indicative of activity on a computer network, by using a computation engine implemented using Apache Storm or Apache Spark Streaming;

generating real-time analysis result data representing the first indicia;

storing the real-time analysis result data in a non-volatile storage system;

retrieving, from the non-volatile storage system, the real-time analysis result data and second event data indicative of activity on the computer network; and

detecting, in a batch mode, second indicia of possible security breaches based on the second event data and the real-time analysis result data, by using an Apache Spark cluster computing engine.

22. The method of claim 21, further comprising:

using outputs of the Apache Spark cluster computing engine, in conjunction with the first event data, to detect the first indicia of possible security breaches.

23. The method of claim 21, wherein using the computation engine includes using 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 using the Apache Spark cluster computing engine includes using 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.

24. The method of claim 21, wherein:

using the computation engine includes executing 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

using the Apache Spark cluster computing engine includes executing 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.

25. The method of claim 21, wherein:

using the computation engine and using the Apache Spark cluster computing engine collectively include using 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.

26. A non-transitory machine-readable storage medium storing instructions, execution of which in a computer system causes the computer system to perform operations comprising:

detecting, in a real-time processing mode, first indicia of possible security breaches based on first event data indicative of activity on a computer network, by executing a computation engine implemented using Apache Storm or Apache Spark Streaming;

generating real-time analysis result data representing the first indicia;

storing the real-time analysis result data in a non-volatile storage system;

retrieving, from the non-volatile storage system, the real-time analysis result data and second event data indicative of activity on the computer network; and

detecting, in a batch mode, second indicia of possible security breaches based on the second event data and the real-time analysis result data, by executing an Apache Spark cluster computing engine.

27. The non-transitory machine-readable storage medium of claim 26, the operations further comprising:

using outputs of the Apache Spark cluster computing engine, in conjunction with the first event data, to detect the first indicia of possible security breaches.

28. The non-transitory machine-readable storage medium of claim 26, such that:

using the computation engine includes using 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

using the Apache Spark cluster computing engine includes using 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.

29. The non-transitory machine-readable storage medium of claim 26, such that:

using the computation engine includes executing 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

using the Apache Spark cluster computing engine includes executing 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.

30. The non-transitory machine-readable storage medium of claim 26, wherein:

using the computation engine and using the Apache Spark cluster computing engine collectively include using 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.