A potentially serious vulnerability that can be exploited for path traversal and arbitrary code execution has been found in Kubernetes, the popular open source container orchestration system.
Last year, developers of Kubernetes and OpenShift, a Red Hat-made container application platform that uses Kubernetes, patched a directory traversal vulnerability (CVE-2018-1002100) that could have been exploited to write arbitrary files to arbitrary paths on the system using malicious containers.
Researchers at container security firm Twistlock discovered that the fix was incomplete and demonstrated that attacks were still possible. A new CVE identifier has been assigned, CVE-2019-1002101, and Kubernetes developers released versions 1.11.9, 1.12.7, 1.13.5, and 1.14.0 to address the problem, which they have classified as “high severity.”
Both the new and the old flaws are related to Kubectl, a command-line interface for running commands against Kubernetes clusters. Specifically, the issue involves the cp command, which is used to copy files between a container and the user’s machine. During this copying process, a tar is created inside the container, it’s copied over the network, and then it’s unpacked by kubectl on the user’s device.
However, if an attacker can plant a malicious tar binary inside the container, they could exploit these flaws to write arbitrary files to any location on the system when the Kubectl cp command is used.
Twistlock researchers explained that an attacker can exploit this directory traversal to steal potentially sensitive information from the devices of Kubectl users, which are typically developers, administrators or members of DevOps teams.
It may also be possible to execute arbitrary code on the system, but that depends on the level of permissions Kubectl has. If it runs as root, an attacker could easily modify one of the configuration or system files that are executed when the system boots up.
Since Kubectl is more often run with user privileges, achieving code execution may not be as easy. However, Twistlock experts have described a simple attack method that could work in many scenarios.
“This vulnerability can be dangerous in one of two scenarios: 1) A user unknowingly downloads a malicious container image with a bad tar. The attacker can push such an image to any registry (e.g. Docker Hub) for a popular image he has control of or rely on typosquatting. 2) An attacker compromises a running container by exploiting another vulnerability or in some cases he may have legitimate access to a container. The attacker then plants a malicious tar replacing the original tar of the image,” said Twistlock researcher Ariel Zelivansky.
“A sophisticated attacker will also conceal the exploit by making the malicious tar include the originally requested files in its output, or make the symlink file a dotfile, so that the attack is not immediately apparent,” Zelivansky added.
Amazon Web Services (AWS) has also released a security bulletin for this and another, less severe vulnerability found recently in Kubernetes.
Related: Critical Privilege Escalation Flaw Patched in Kubernetes
Related: Misconfiguration a Top Security Concern for Containers
Related: Researchers Find 21,000 Exposed Container Orchestration Systems