IPMI 2.0 RAKP Password Hash Disclosure
The IPMI 2.0 RAKP handshake returns a salted password hash for any valid BMC user before authentication, which can be cracked offline to recover the password.
| Severity | High |
| CVSS 3.1 | 8.1 CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H Open in the calculator |
| CVSS 4.0 | 9.2 CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N Open in the calculator |
| CWE | CWE-201: Insertion of Sensitive Information Into Sent Data |
| OWASP Top 10 | A01:2025 Broken Access Control |
| MITRE ATT&CK | T1110.002 Password Cracking, T1078 Valid Accounts |
| CVE | CVE-2013-4786 |
| Category | Information Disclosure |
Description
The baseboard management controller (BMC) supports IPMI 2.0 over the network, whose RMCP+ Authenticated Key-Exchange Protocol (RAKP) sends the client an HMAC computed from the requested user's password before the client has proved who it is. Any party that can reach the BMC can therefore request this value for a valid user name and try to recover the password offline. This is a weakness in the IPMI 2.0 specification itself (CVE-2013-4786) rather than a fault in one vendor's implementation, so every BMC with IPMI over LAN enabled is affected.
The captured values can be cracked with tools such as Hashcat, and weak or vendor-default BMC passwords are usually recovered quickly, while long, randomly generated BMC passwords make cracking impractical and are the main factor that limits this issue. An attacker who recovers a BMC password can log in to the BMC with that account's privileges, which commonly include administration. That gives control of the server through the BMC: power control, the remote console, virtual media and BMC user management, and from there a route to the operating system and its data. The same password is often reused across a fleet of BMCs or for other systems, which widens the impact.
The issue is identified on UDP 623, for example with the Metasploit ipmi_dumphashes module, which requests hashes for a list of common user names and reports those that are valid, because the BMC answers differently for user names that do not exist. Scanners report the issue as IPMI 2.0 RAKP remote password hash retrieval.
Remediation
The weakness is part of the IPMI 2.0 protocol and cannot be configured away while IPMI over LAN is enabled, so reduce exposure and make the hashes useless:
- Disable IPMI over LAN on BMCs where it is not needed, and manage them through the vendor's web interface or Redfish API over HTTPS instead.
- Where IPMI over LAN is required, restrict UDP 623 to a dedicated management network that only administrators and management systems can reach.
- Set long, randomly generated, unique passwords on every BMC account, kept in a password manager, and change any password that may have been captured or is a vendor default.
- Rename or remove default accounts where the platform allows, and disable unused and anonymous accounts, so that valid user names are harder to guess.
Confirm the fix by repeating the hash retrieval test, checking that UDP 623 cannot be reached from user networks, and checking that any hashes still obtainable from the management network cannot be cracked with common wordlists.
References
Add This Finding to Your Library
In Pental this write-up is in Settings, Library. Add to my library copies it into your own, rated in your firm's CVSS version, ready to put in your own words.
Free trialYour trial starts when your portal goes live, not when you sign up, so the time you spend setting up does not count. · No card required
Only need the reporting? Our reporting tool is free and open source, with the same library.