OpenVPN is cryptographically well-regarded, so attacking it means attacking how it is deployed rather than the protocol. The richest target is the client profile (.ovpn): it often embeds the CA, client certificate, private key, and sometimes credentials, so a stolen profile authenticates the attacker directly. The optional tls-auth/tls-crypt HMAC protects the control channel; when it is absent or its key is leaked (it too lives in the profile), the control channel is exposed to probing and some attacks. Where OpenVPN uses username/password auth, the server is a spraying target. And static-key or certificate reuse across clients means one recovered key opens the VPN.
# steal/examine a client profile for embedded keys and creds
grep -iE '<cert>|<key>|<tls-auth>|auth-user-pass' client.ovpn
openssl rsa -in client.key -noout -text 2>/dev/null # confirm a usable private key
# reuse the profile to connect as that client
openvpn --config client.ovpn
# where username/password auth is used, spray the server (see password brute force)
Exploitation notes#
- Client profiles are the prize: hunt for
.ovpnfiles (and.key/.crt) on endpoints, backups, images, and repositories; an embedded private key and certificate authenticate as that client with no further credential. - A missing
tls-auth/tls-crypt, or a leaked HMAC key (recoverable from a profile), removes the control-channel protection that otherwise blocks unauthenticated probing of the OpenVPN server. - Shared/static keys and reused client certificates mean one recovered credential opens the VPN for the attacker; check whether the deployment issues per-client certs or reuses one.
- Username/password OpenVPN is sprayable like any portal (password brute force); certificate reuse and theft follow certificate abuse.