wiki:CCSInjection

OpenSSL CCS Injection Vulnerability (CVE-2014-0224)

On June 5th, 2014, a number of vulnerabilities in OpenSSL were disclosed (see https://www.openssl.org/news/secadv_20140605.txt). One of those, the CCS Injection Vulnerability, affects OpenVPN. This page discusses the consequences for OpenVPN users.

What does the CCS Injection Vulnerability mean?

In short: if both the client and the server are running a vulnerable version of OpenSSL, an active attacker with a man-in-the-middle position can trick OpenSSL to use keys known to the attacker. This means the attacker can read and even manipulate everything on the TLS connection. In the OpenVPN case, that includes the traffic protection keys for your VPN data, and thus your VPN data. For more information, visit the CCS Injection Vulnerability page at http://ccsinjection.lepidum.co.jp/ or check the CVE at http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0224.

Use of TLS auth prevents this vulnerability from being exploited.

What should I do?

  • Update your OpenSSL and restart OpenVPN (and any other daemons using it).
  • If you're using the OpenVPN Windows installer upgrade to the latest OpenVPN release.
  • Use TLS-auth as an extra layer of protection (see Hardening).

Do I need to create new private keys and certificates?

No, those are not leaked by this vulnerability. The keys an attacker could intercept are the temporary TLS and VPN data channel keys, which are freshly generated for each new connection.

Do the six other OpenSSL vulnerabilities affect OpenVPN?

No. Current OpenVPN releases (that is, upto 2.3.4) do not use DTLS, SSL_MODE_RELEASE_BUFFERS or ECDH, and are thus not affected by bugs in those components.

Last modified 5 years ago Last modified on 06/05/14 18:58:08