services / Google Cloud / Kubernetes Deployments

Control Kubernetes Deployment objects in a given cluster.

Deployments are declarative updates to Kubernetes Pods and ReplicaSets. They provide an abstraction layer that allows services to discover and connect to pods running within a Deployment. Applications can scale, load balance, and seamlessly handle changes in the underlying pod instances. Access to Deployments is sensitive because they are a primary interface for controlling applications running on Kubernetes. A Deployment specifies the container image to run, and may lead to arbitrary code execution in the cluster. A good mitigation strategy is to make Kubernetes clusters private, which effectively disables pulling images over the internet. Binary Authorization in GKE is an additional layer of security to defend against supply chain attacks. Secondly, creating or updating the replica count of Deployments drains the limited resources available to other Kubernetes workloads.

container.​deployments.​create

Deployments tie up compute resources in Kubernetes that cannot be allocated to another Kubernetes workload while the deployment is present. Deployments run a user-specified container image, which may allow an attacker to escalate their privileges by running arbitrary code inside the Kubernetes workload. Secondly, creating Deployments drains the limited resources available to other Kubernetes workloads.

Risks

1. Data exfiltration (CRITICAL)
2. Lateral movement (BOOST)
3. Network movement (BOOST)
4. Resource hijacking (MEDIUM)
5. Spend (MEDIUM)

Scope: CRITICAL

This privilege may grant access to sensitive data from a significant fraction of organizational functions, allow interruption of critical organizational services, or its exploit could lead to significant privilege escalation.

Links