Skip to content

Missing TLS certificate verification

High severity GitHub Reviewed Published Jul 31, 2020 in faye/faye • Updated May 25, 2023

Package

bundler faye (RubyGems)

Affected versions

< 1.4.0

Patched versions

1.4.0

Description

Faye uses em-http-request and faye-websocket in the Ruby version of its client. Those libraries both use the EM::Connection#start_tls method in EventMachine to implement the TLS handshake whenever a wss: URL is used for the connection. This method does not implement certificate verification by default, meaning that it does not check that the server presents a valid and trusted TLS certificate for the expected hostname. That means that any https: or wss: connection made using these libraries is vulnerable to a man-in-the-middle attack, since it does not confirm the identity of the server it is connected to.

The first request a Faye client makes is always sent via normal HTTP, but later messages may be sent via WebSocket. Therefore it is vulnerable to the same problem that these underlying libraries are, and we needed both libraries to support TLS verification before Faye could claim to do the same. Your client would still be insecure if its initial HTTPS request was verified, but later WebSocket connections were not.

This has been a requested feature in EventMachine for many years now; see for example #275, #378, and #814. In June 2020, em-http-request published an advisory related to this problem and fixed it by implementing TLS verification in their own codebase; although EventMachine does not implement certificate verification itself, it provides an extension point for
the caller to implement it, called ssl_verify_peer. Based on this implementation, we have incorporated similar functionality into faye-websocket.

After implementing verification in v1.1.6, em-http-request has elected to leave the :verify_peer option switched off by default. We have decided to enable this option by default in Faye, but are publishing a minor release with added functionality for configuring it. We are mindful of the fact that this may break existing programs, but we consider it much more important that all clients have TLS verification turned on by default. A client that is not carrying out verification is either:

  • talking to the expected server, and will not break under this change
  • being attacked, and would benefit from being alerted to this fact
  • deliberately talking to a server that would be rejected by verification

The latter case includes situations like talking to a non-public server using a self-signed certificate. We consider this use case to be "working by accident", rather than functionality that was actively supported, and it should be properly and explicitly supported instead.

We are releasing Faye v1.4.0, which enables verification by default and provides a way to opt out of it:

client = Faye::Client.new('https://example.com/', tls: { verify_peer: false })

Unfortunately we can't offer an equivalent of the :root_cert_file option that has been added to faye-websocket, because em-http-request does not support it. If you need to talk to servers whose certificates are not recognised by your default root certificates, then you need to add its certificate (or another one that can verify it) to your system's root set.

The same functionality is now supported in the Node.js version, with a tls option whose values will be passed to the https and tls modules as appropriate when making connections. For example, you can provide your own CA certificate:

var client = new faye.Client('https://example.com/', {
  tls: {
    ca: fs.readFileSync('path/to/certificate.pem')
  }
});

For further background information on this issue, please see faye#524 and faye-websocket#129. We would like to thank Tero Marttila and Daniel Morsing for providing invaluable assistance and feedback on this issue.

References

@jcoglan jcoglan published to faye/faye Jul 31, 2020
Reviewed Jul 31, 2020
Published to the GitHub Advisory Database Jul 31, 2020
Last updated May 25, 2023

Severity

High

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Network
Attack complexity
High
Privileges required
None
User interaction
Required
Scope
Changed
Confidentiality
High
Integrity
High
Availability
None

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:N

EPSS score

0.204%
(59th percentile)

Weaknesses

CVE ID

CVE-2020-15134

GHSA ID

GHSA-3q49-h8f9-9fr9

Source code

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.