Cloud Foundry Cloud Controller (CAPI), versions prior to 1.91.0, logs properties of background jobs when they are run, which may include sensitive information such as credentials if provided to the job. A malicious user with access to those logs may gain unauthorized access to resources protected by such credentials.
cf-deployment
Vendor: cloudfoundry
Security Vulnerability Index
Page 3 / 6Cloud Foundry Cloud Controller API (CAPI), version 1.88.0, allows space developers to list all global service brokers, including service broker URLs and GUIDs, which should only be accessible to admins.
Cloud Foundry UAA Release, versions prior to v74.10.0, when set to logging level DEBUG, logs client_secret credentials when sent as a query parameter. A remote authenticated malicious user could gain access to user credentials via the uaa.log file if authentication is provided via query parameters.
Cloud Foundry UAA Release, versions prior to v74.8.0, logs all query parameters to tomcat’s access file. If the query parameters are used to provide authentication, ie. credentials, then they will be logged as well.
Cloud Foundry Routing, all versions before 0.193.0, does not properly validate nonce input. A remote unauthenticated malicious user could forge an HTTP route service request using an invalid nonce that will cause the Gorouter to crash.
Cloud Foundry SMB Volume, versions prior to v2.0.3, accidentally outputs sensitive information to the logs. A remote user with access to the SMB Volume logs can discover the username and password for volumes that have been recently created, allowing the user to take control of the SMB Volume.
Cloud Foundry UAA, versions prior to v74.3.0, contains an endpoint that is vulnerable to SCIM injection attack. A remote authenticated malicious user with scim.invite scope can craft a request with malicious content which can leak information about users of the UAA.
Cloud Foundry NFS Volume Service, 1.7.x versions prior to 1.7.11 and 2.x versions prior to 2.3.0, is vulnerable to LDAP injection. A remote authenticated malicious space developer can potentially inject LDAP filters via service instance creation, facilitating the malicious space developer to deny service or perform a dictionary attack.
Cloud Foundry cf-deployment, versions prior to 7.9.0, contain java components that are using an insecure protocol to fetch dependencies when building. A remote unauthenticated malicious attacker could hijack the DNS entry for the dependency, and inject malicious code into the component.
Cloud Foundry Diego, release versions prior to 2.8.0, does not properly sanitize file paths in tar and zip files headers. A remote attacker with CF admin privileges can upload a malicious buildpack that will allow a complete takeover of a Diego Cell VM and access to all apps running on that Diego Cell.