Join our Newsletter — 33% off our NHI Course
Home› Glossary› Cyber Security› Rolling Update
Cyber Security

Rolling Update

← Back to Glossary
By NHI Mgmt Group Updated October 31, 2026 Domain: Cyber Security

A rolling update is a deployment method that replaces application instances gradually rather than all at once. In Kubernetes, this helps teams introduce new versions while keeping the service available and limiting disruption. It is commonly used when change risk must be controlled and downtime is not acceptable.

What a rolling update means operationally

A rolling update is a release pattern, not just a deployment command. It changes application instances in stages so traffic can continue flowing while the system is gradually replaced. That makes it a common choice for services that need availability during change, especially when the new version must prove itself before the old one is fully retired.

The core trade-off is control versus speed. A rolling update reduces the blast radius of a bad release, but it also means old and new versions may run side by side for a time. Teams therefore need to think about version compatibility, traffic routing, and whether mixed-version behaviour is acceptable while the rollout is in progress.

Why rolling updates are used in Kubernetes

In Kubernetes, rolling updates are the default way to replace Pods without dropping the whole workload at once. The scheduler and controller can bring up new Pods, wait for them to become ready, and then scale down older Pods in sequence. This is what lets the service stay online while the deployment progresses.

The method is especially useful for stateless services, horizontally scaled applications, and environments where even short outages are costly. It also supports safer delivery because readiness gates can prevent traffic from reaching instances that are not yet prepared to serve requests. That makes the deployment process more resilient than a big-bang replacement.

What can go wrong during a rolling update

Rolling updates are safer than an all-at-once cutover, but they are not risk-free. Problems often come from version skew, where requests land on both old and new instances at the same time, or from a bad readiness signal that sends traffic to a Pod before it is truly healthy. If the new build has a latent defect, the rollout can propagate the problem gradually across the fleet.

They also interact with application state and dependency timing. A database change, API contract change, or cache assumption that is not backward compatible can make the rollout fail even though the deployment mechanism itself is working correctly. In practice, the deployment pattern is only as safe as the release discipline behind it.

How rolling updates fit into release reliability

Rolling updates are one part of a broader reliability model. They help limit disruption, but they do not replace health checks, rollback planning, or compatibility testing. The technique works best when teams can observe the rollout in real time and stop or reverse it if error rates, latency, or failed readiness checks start to rise.

They also create a temporary state where two versions coexist, which is useful for resilience but awkward for defects that depend on uniformity. That is why release engineering often pairs rolling updates with feature flags, canary-style observation, and strict interface discipline. The deployment method reduces operational risk, but it does not eliminate release risk.

Risk and Threat Considerations

Rolling updates can widen exposure if a flawed build is deployed gradually across many instances before the problem is noticed. They also increase the importance of deployment trust, because a compromised or defective package can persist long enough to affect live traffic while the old version is still being retired.

Failure mechanism: A bad readiness check, incompatible schema change, or maliciously altered release can move from a small subset of instances to the broader fleet before operators detect the issue.

Impact: The result can be partial outage, inconsistent user experience, data or request failures, and a slower containment path than a clean rollback from a single cutover.

Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

NIST SP 800-53 Rev 5, NIST CSF 2.0 and CIS Controls v8 set the technical controls, while ISO/IEC 27001:2022 defines the regulatory obligations.

FrameworkControl / ReferenceRelevance
NIST SP 800-53 Rev 5CM-3 — Configuration Change ControlRolling updates are a controlled production change process.
SI-2 — Flaw RemediationA rolling update is commonly used to deploy fixes safely across live instances.
Recommendation — Apply CM-3 to review, approve, and track release changes before gradual production rollout. Use SI-2 to validate and deploy fixes, then confirm affected instances are remediated.
NIST CSF 2.0PR.IR-02 — Patches and updates are applied to systemsThe term describes a method for applying updates while maintaining service continuity.
Recommendation — Use PR.IR-02 to maintain patching and update processes that preserve availability during change.
CIS Controls v8CIS-7 — Continuous Vulnerability ManagementRolling updates often support staged remediation and controlled exposure reduction.
Recommendation — Use CIS-7 to prioritise, stage, and verify remediation across production assets.
ISO/IEC 27001:2022A.8.32 — Change managementRolling updates are a production change method that needs controlled release governance.
Recommendation — Apply A.8.32 to govern release changes, approvals, and rollback expectations.

Practitioner Guidance

What to watch for: Treat a rolling update as a controlled exposure window, not a guarantee of safety. The rollout should be observable enough that readiness, error rates, and version skew can be evaluated while the change is still in progress.

Practitioner note: The strongest deployments pair gradual replacement with explicit rollback criteria and compatibility checks, because the deployment strategy only works when the application can tolerate mixed versions for a short period.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Reviewed and updated by the NHIMG editorial team on October 31, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org