Domain Fronting

Created the Sunday 15 January 2023. Updated 6 months, 3 weeks ago.

Domain fronting is a technique used to hide the true destination of internet traffic. It may be used to hide the real destination of an offensive payload or is could be used to bypass censorship or any form of network restriction. The technique may be used through CDN infrastructure or on any infrastructure that perform separation between SSL and http backend.

It involves SSL/TLS connection since it abuse of the destination requested in the SNI (Server Name Indication) field at SSL level at "Front" and the real hosname requested by the HOST header at the HTTP level. In clear you declare one destination at SSL level and another one at HTTP level. From an defender point on view you see only the SSL destination in the firewall logs and not the real final destination.

Additionally, some other headers may be used to facilitate domain fronting like "X-Forwarded-Host" and "X-Originating-URL". But these headers may vary depending on the implementation.

Such techniques were exploited by groups like APT29. Hovewer, some cloud providers like AWS, Google Cloud, and Cloudflare among others, have started to detect and block domain fronting attempts. Since then a new technics named domain hiding as been discovered. This is a complementary technique that as been presented by Erik Hunstad at defcon 2020. This techniques use the ESNI header in the TLS3.1 version.

Technique Identifiers

U0908 T1090.004

Technique Tags

Domain fronting bypass censorship network restriction SSL/TLS connection SNI (Server Name Indication) field X-Forwarded-Host X-Originating-URL APT29 TLS3.1 version

Code Snippets

import requests
import re

# The actual destination domain requested
destination_domain = "www.youtube.com"

# The fronting domain used in ssl
fronting_domain = "www.google.com"

# Create a session object
session = requests.Session()

# force real destination in the Host header
session.headers["Host"] = destination_domain

# Send the request to the fronting domain
response = session.get("https://" + fronting_domain, headers={'X-Originating-URL': 'https://' + destination_domain})

# Extract the title of the received page
title = re.search("<title>(.*)</title>", response.text)
print(title. Group(1))

Additional Resources

External Links

The resources provided below are associated links that will give you even more detailed information and research on current evasion technique. It is important to note that, while these resources may be helpful, it is important to exercise caution when following external links. As always, be careful when clicking on links from unknown sources, as they may lead to malicious content.