Search for Files on Machines

Note

To use these features, you can add the DFIR package to your instance of the Cybereason platform for an additional cost. Contact your Customer Success representative to request access to this package.

When investigating malicious operations, you can search for problematic files in the following scenarios:

  • You have intelligence from global threat intelligence feeds about a malicious file.

  • You found a malicious file on a machine in your organization and want to see if other machines contain the same file.

  • You know the file is malicious but the file has not run on a machine. Because the file has not run, the Cybereason platform may not have detected the file as part of the sensor’s normal data collection.

To address these situations, you can use the Cybereason platform’s Live File search screen to search for malicious files throughout your organization’s machines.

You can also search for files using the Cybereason API. For details see the Hunting and Investigation API Reference section of the Cybereason API Reference.

Perform a file search with YARA rules

You can search for suspicious files using YARA rules. You can upload a YAR file and filter according to the rules in that file. You can also use YARA rule searches together with the other filters on the File search screen.

Note

File search with YARA rules is not supported for machines running supported versions of Linux.

You use YARA rules to identify:

  • Malware

  • IOCs of malware

  • Any suspicious files that could be a sign of an attack an analyst would like to identify

YARA rules enable you to use binary patterns to identify malware. This is helpful in cases where a malware file was slightly tampered with (e.g. its hash was altered) compared with the file known to threat intelligence sources, making it hard to detect using other methods.

To search with YARA rules, follow these steps:

  1. In the File search screen, select the YARA search tab.

    YARA rules search in the File search screen

  2. In the YARA search tab, in the YAR file option section, click Upload YAR file to upload a YAR file containing the YARA rules to use in the search. The YAR file size must be less than 5 MB.

  3. If needed, for the Time option, select Created or Modified, an operator (Before, After, or Between) and then enter a time from the calendar widget.

  4. In the Size field, select an operator (Greater than, Less than, Equals, or Between), a file size unit (Bytes, KB, MB, or GB) and a file size.

  5. In the Folders option, add at least one folder path.

    The folder path is necessary since otherwise, the search would take too much time and resources. A warning is displayed if no folder is enter and the Search button is disabled without entering a folder path.

  6. In the Location section, create a filter to limit the search to specific machines.

    Possible filters for machines for file search

    Cybereason recommends that you use this field to limit the file search effect on your servers. If you do not set a filter for machines for a file search operation, the file search operation runs on all machines currently connected to your Cybereason platform.

    In the Machines field, you can specify machines in a number of ways:

    Search all machines

    In the dropdown list, select All.

    If you select this option, there is a maximum of 5000 machines for the file search operation.

    Search machines that meet a specific filter criteria

    In the dropdown list, select Query.

    You should use this option when you know the parameters of a file search operation, such as machines of a certain operating system, online machines, and so forth.

    This option enables you to search any machine based on machine characteristics. Available characteristics for filtering include:

    • Machine name

    • Machine FQDN

    • Sensor version

    • CPU usage

    • Memory usage

    • OS

    • OS version

    • Anti-Ransomware mode

    • PowerShell mode

    • Remote Shell mode

    • Data collection mode

    • Internal IP address of the machine

    • External IP address of the machine

    • Isolated status of the machine

    • App control mode

    • Sensor status

    • Service status

    • Outdated status of the machine

    • Signatures mode for the Anti-Malware > Signatures mode

    • Signatures DB version

    • Anti-Malware mode

    • Exploit Protection mode

    • Behavioral doc mode

    • Behavioral doc sensitivity

    • AI Detect mode

    • AI Prevent mode

    • Signatures mode origin

    For a more detailed description of these options, see Export Sensor Metadata.

    Search a specific list of machines

    In the dropdown list, select CSV file.

    This option is best used when there were failures on a previous file search and you need to run the file search operation on specific machines.

    Click Upload CSV with machine list* to add the CSV file with the machine names.

    In the CSV file, you must add a column named MachineName with a list of machine names.

    For example, you could prepare a CSV file like this:

    MachineName

    Win10-x64-20h1

    Win10-x64-rs5

    ron10x64

    win10-x64-20h1

    By default, the search timeout is 10 minutes. You can select any interval from five minutes to sixty minutes and click Apply.

  7. Click Search.

    While the Cybereason platform scans files against the YARA rules, files are temporarily locked for a few seconds, and end users cannot delete the file.

    For large files, only the first 100 MB of each file is scanned.

    If a file takes over 10 seconds to scan, the file search operation times out and proceeds to the next file.

    Note

    If your environment uses sensor grouping and you have the Local Responder role assigned for your Cybereason user, you can view only the results from sensors in the groups to which you are assigned.

In search results, the Matched YARA rules column displays the YARA rules that matched the rules in the YAR file:

YARA rules

Note

If many sensors seem to have stopped, check the YARA rules you entered. If they are too complex, sensors may stop if they experience heavy load when scanning using these rules.