The core foundation of GitOps is that Git serves as the single source of truth for the desired state of both applications and infrastructure. This means all configuration is declared in Git in a version-controlled, auditable, and verifiable manner. Operators and reconciliation agents continuously pull these definitions to ensure the live system matches what Git declares.

“Declarative descriptions of the desired state of the system must be versioned in Git, making Git the single source of truth. This provides auditability, reliability, and enables rollbacks by reverting changes in Git.”

This principle ensures that any change in system state is traceable through Git commits, making environments predictable, reproducible, and transparent.

Progressive delivery is a GitOps pattern used to release software gradually, reducing risks associated with deploying new versions. Techniques such as canary releases, feature flags, and blue-green deployments allow teams to incrementally roll out changes, validate functionality with subsets of users, and minimize potential disruptions.

“Progressive delivery builds on continuous delivery by enabling safer, incremental rollouts. This pattern reduces risk, improves reliability, enhances user experience, and allows for validation of features with a portion of users before wider release.”

Therefore, progressive delivery is useful in multiple scenarios (not just one-time deployments or small teams), making option C correct.

In GitOps, Git is the single source of truth . If changes are made manually in the cluster (via kubectl ), those changes will drift from the desired state in Git. To ensure consistency, the correct approach is to update the Git repository (Desired State store) so that the reconciler can continuously apply and maintain those changes.

“The desired state must always be declared in Git. Manual changes in the cluster will be overwritten by reconciliation unless they are committed to the Git repository.”

Thus, the correct answer is B .

GitOps requires a pull-based approach as the default model, where agents in the cluster automatically pull desired state from Git. However, push-based triggers (e.g., webhooks) can optionally be used to complement pull-based reconciliation.

“Pull-based reconciliation is required in GitOps to ensure automation and security. Push-based triggers may optionally complement this, but the pull model is fundamental.”

Thus, the correct answer is D .

The State Store is typically implemented using a version control system (VCS) such as Git. This ensures that the system’s state is auditable, versioned, and immutable. Other systems like databases or distributed file systems do not meet the immutability and versioning requirements of GitOps.

“In GitOps, the desired state must be stored in a version control system. This serves as the State Store, providing a complete, immutable, and auditable history of changes.”

Thus, the correct answer is D: version control system .

A GitOps-managed Kubernetes cluster uses reconciliation loops to continuously compare the actual state of the system with the desired state declared in Git. When drift (divergence between declared configuration and live cluster state) is detected, the GitOps operator automatically reconciles the difference to bring the system back into alignment.

“In GitOps, a reconciliation loop ensures that the desired state as declared in Git is continuously compared with the observed state of the system. If drift is detected, the system automatically takes corrective action to reconcile the difference and restore the declared configuration.”

This ensures consistency, reliability, and self-healing. Manual intervention is not required for drift correction, as the automated reconciliation is a core principle of GitOps.

GitOps emphasizes immutability and auditability. If an error occurs in the configuration stored in Git, the system should be rolled back to the last known good state while the error is fixed. This preserves system reliability and aligns with the GitOps principle of rollback through version control.

“With Git as the source of truth, if an error is introduced, the system can be rolled back by reverting to a previous commit. This ensures stability while the faulty configuration is corrected.”

Thus, the recommended action is C: Roll back to the previous working version .