KCSA · Domain 1: Overview of Cloud Native Security · 14% of the exam
Task 1.3: Controls and Frameworks
Preventive, detective and corrective controls, the Kubernetes security checklist, and how benchmarks and frameworks shape a cluster baseline.
Study it
Controls, the security checklist and benchmarks
Lesson coming
Sample questions
Try each one before opening the answer. Every option is explained, with the CNCF documentation page that proves it.
Question 1 · choose 1
Offboarding at a company disables a leaver's account in its identity provider, but RoleBindings that name the user are left in the clusters. The identity provider later reissues freed usernames to new hires. Which control from the Kubernetes RBAC good practices addresses the risk this creates?
AShorter lifetimes for the identity provider's login sessions
BRecording every change to RBAC objects in the API server audit log
CEnforcing the Restricted Pod Security Standard in every namespace users can reach
DPeriodic RBAC reviews that remove entries for users who no longer exist
Show the answer and why
AShorter lifetimes for the identity provider's login sessions
Incorrect
Sessions do end sooner, but a new hire who signs in with the reused name gets a fresh session and still matches the old bindings.
BRecording every change to RBAC objects in the API server audit log
Incorrect
An audit log shows who changed bindings, but the stale bindings were never changed; they remain in place and keep granting rights to the reused name.
CEnforcing the Restricted Pod Security Standard in every namespace users can reach
Incorrect
Pod Security Standards limit Pod settings. They do not change which API permissions a binding grants to a username.
DPeriodic RBAC reviews that remove entries for users who no longer exist
Correct
The good practices call for periodic RBAC reviews because an account created with the same name as a deleted user automatically inherits that user's rights.
Periodic review is a control the RBAC good practices name explicitly: redundant entries must be removed because a newly created account with a deleted user's name inherits every right still bound to that name.
The Kubernetes security checklist sets a limit on how far in the future intermediate and leaf certificates may expire. What is it?
ANo more than 90 days
BNo more than 10 years
CNo limit, as long as the root certificate is offline
DNo more than 3 years
Show the answer and why
ANo more than 90 days
Incorrect
The checklist does not set a 90-day limit for these certificates.
BNo more than 10 years
Incorrect
Ten years is far beyond the checklist's limit of three years.
CNo limit, as long as the root certificate is offline
Incorrect
Protecting the root certificate is a separate checklist item; the expiry limit still applies.
DNo more than 3 years
Correct
The checklist item reads that intermediate and leaf certificates have an expiry date no more than 3 years in the future.
Under Authentication and Authorization, the security checklist asks for a protected root certificate and for intermediate and leaf certificates that expire no more than three years in the future.
How does the Kubernetes security checklist say the cluster's root certificate should be protected?
ABy storing its private key in a Secret in kube-system for the controller manager
BBy keeping it in an offline CA, or a managed online CA with effective access controls
CBy copying it to every node for redundancy
DBy publishing it so that clients can verify it
Show the answer and why
ABy storing its private key in a Secret in kube-system for the controller manager
Incorrect
A Secret is readable by anyone allowed to read it and by etcd's operators; the checklist asks for an offline or access-controlled CA.
BBy keeping it in an offline CA, or a managed online CA with effective access controls
Correct
The checklist item says the root certificate is protected either as an offline CA or as a managed online CA with effective access controls.
CBy copying it to every node for redundancy
Incorrect
Spreading the root key widens its exposure, the opposite of protecting it.
DBy publishing it so that clients can verify it
Incorrect
Publishing the CA certificate is normal, but the checklist is about protecting the root CA itself.
The root certificate anchors trust for the cluster's PKI. The security checklist asks for it to be kept offline or in a managed online CA with effective access controls.