Описание
Lemur subject to insecure random generation
Overview
Lemur was using insecure random generation for its example configuration file, as well as for some utilities.
Impact
The potentially affected generated items include:
| Configuration item | Config option name (if applicable) | Documentation link (if applicable) | Rotation option | Code reference(s) |
|---|---|---|---|---|
| Flask session secret | SECRET_KEY | Flask documentation | Generate a new secret and place in config; all existing sessions will be invalidated | N/A, internal to Flask |
| Lemur token secret | LEMUR_TOKEN_SECRET | Lemur's configuration documentation | Generate a new secret and place in config; all existing JWTs will be invalidated and must be regenerated (including API keys) | 1, 2 |
| Lemur database encryption key | LEMUR_ENCRYPTION_KEYS | Lemur's configuration documentation | A new key can be generated and added to this list, but existing data encrypted with prior keys cannot be re-encrypted unless you write a custom re-encryption process | 1 |
| OAuth2 state token secret key | OAUTH_STATE_TOKEN_SECRET | Lemur's configuration documentation | Generate a new secret and place in config | 1 |
| Randomly generated passphrases for openssl keystores | N/A, generated at runtime but persisted | N/A | Re-export all openssl keystores and replace them wherever they're in use | 1 |
| Initial password for LDAP users | N/A, generated at runtime but persisted | N/A | N/A, cannot be rotated* | 1 |
| Initial password for Ping/OAuth2 users | N/A, generated at runtime but persisted | N/A | N/A, cannot be rotated* | 1 |
| Oauth2 nonce | N/A, short-lived runtime secret | N/A | N/A, rotation is not required (these are short-lived) | 1 |
| Verisign certificate enrollment challenges | N/A, short-lived runtime secret | N/A | N/A, rotation is not required (these are short-lived) | 1 |
If your deployment of Lemur is using any of the above config secrets that were generated by Lemur's example config (i.e., generated using insecure randomness), you should rotate those config secrets. If you generated your config secrets in a more secure way, they are not known to be compromised, but you should still upgrade Lemur to ensure that you receive code fixes for the runtime-generated secrets.
For general information and guidance on Lemur secret configuration, see Lemur's configuration documentation, which includes information on many of the configuration options listed above.
*For the user passwords: Even though these users are configured to use SSO, they do get generated with valid database passwords that can be used to log in. Since Lemur doesn't have an option to change passwords (#3888), one option for rotating them would be to directly modify the value in the database to some other unguessable string (you do not need to know the plaintext password since it won't be used).
Patches
The patch is available in v1.3.2.
Workarounds
No workarounds are available.
References
N/A
Ссылки
- https://github.com/Netflix/lemur/security/advisories/GHSA-5fqv-mpj8-h7gm
- https://nvd.nist.gov/vuln/detail/CVE-2023-30797
- https://github.com/Netflix/lemur/issues/3888
- https://github.com/Netflix/lemur/commit/666d853212174ee7f4e6f8b3b4b389ede1872238
- https://github.com/Netflix/security-bulletins/blob/master/advisories/nflx-2023-001.md
- https://github.com/pypa/advisory-database/tree/main/vulns/lemur/PYSEC-2023-20.yaml
- https://vulncheck.com/advisories/netflix-lemur-weak-rng
Пакеты
lemur
< 1.3.2
1.3.2
Связанные уязвимости
Netflix Lemur before version 1.3.2 used insufficiently random values when generating default credentials. The insufficiently random values may allow an attacker to guess the credentials and gain access to resources managed by Lemur.