Detection: DNS Query Length Outliers - MLTK

Description

The following analytic identifies DNS requests with unusually large query lengths for the record type being requested. It leverages the Network_Resolution data model and applies a machine learning model to detect outliers in DNS query lengths. This activity is significant because unusually large DNS queries can indicate data exfiltration or command-and-control communication attempts. If confirmed malicious, this activity could allow attackers to exfiltrate sensitive data or maintain persistent communication channels with compromised systems.

Search

 1
 2| tstats `security_content_summariesonly` count min(_time) as start_time max(_time) as end_time values(DNS.src) as src values(DNS.dest) as dest from datamodel=Network_Resolution by DNS.query DNS.record_type 
 3| search DNS.record_type=* 
 4|  `drop_dm_object_name(DNS)` 
 5| `security_content_ctime(firstTime)` 
 6| `security_content_ctime(lastTime)` 
 7| eval query_length = len(query) 
 8| apply dns_query_pdfmodel threshold=0.01 
 9| rename "IsOutlier(query_length)" as isOutlier 
10| search isOutlier > 0 
11| sort -query_length 
12| table start_time end_time query record_type count src dest query_length 
13| `dns_query_length_outliers___mltk_filter`

Data Source

No data sources specified for this detection.

Macros Used

Annotations

Default Configuration

This detection is configured by default in Splunk Enterprise Security to run with the following settings:

Implementation

To successfully implement this search, you will need to ensure that DNS data is populating the Network_Resolution data model. In addition, the Machine Learning Toolkit (MLTK) version 4.2 or greater must be installed on your search heads, along with any required dependencies. Finally, the support search "Baseline of DNS Query Length - MLTK" must be executed before this detection search, because it builds a machine-learning (ML) model over the historical data used by this search. It is important that this search is run in the same app context as the associated support search, so that the model created by the support search is available for use. You should periodically re-run the support search to rebuild the model with the latest data available in your environment. This search produces fields (query,query_length,count) that are not yet supported by ES Incident Review and therefore cannot be viewed when a notable event is raised. These fields contribute additional context to the notable. To see the additional metadata, add the following fields, if not already present, to Incident Review - Event Attributes (Configure > Incident Management > Incident Review Settings > Add New Entry):

• Label: DNS Query, Field: query
• Label: DNS Query Length, Field: query_length
• Label: Number of events, Field: count Detailed documentation on how to create a new field within Incident Review may be found here: https://docs.splunk.com/Documentation/ES/5.3.0/Admin/Customizenotables#Add_a_field_to_the_notable_event_details

Known False Positives

If you are seeing more results than desired, you may consider reducing the value for threshold in the search. You should also periodically re-run the support search to re-build the ML model on the latest data.

Associated Analytic Story

• Command And Control

• Hidden Cobra Malware

• Suspicious DNS Traffic

Risk Based Analytics (RBA)

Detection Testing

Replay any dataset to Splunk Enterprise by using our replay.py tool or the UI. Alternatively you can replay a dataset into a Splunk Attack Range

Source: GitHub | Version: 4