Rollback
Rollback is the reversal or discarding of changes to restore data, software, or system state to an earlier or otherwise valid state.
Machine Learning Platform Engineering
A rollback is the restoration of a deployed application to a previously functioning version.
- How it is used
- Protects production availability by restoring a functioning application version after a problematic Deployment update.
- How deep it goes
- Explains rollback through rolling updates, ReplicaSets, status checks, events, and a kubectl command, without detailing revision internals.
- Examples
- Uses a hello-joker Deployment, v1-to-v2 image update, rollout events, status output, and kubectl rollout undo.
- What you can do after
- Operate/configure Kubernetes Deployments by applying image updates, inspecting rollout status, and undoing failed releases.
Read how Machine Learning Platform Engineering explains rollback →
.NET in Action, Second Edition
A rollback discards the changes made within a database transaction.
- How it is used
- Prevents a failed money transfer from leaving only one account balance updated.
- How deep it goes
- Works through transaction creation, commit, exception handling, rollback, and change-tracker cleanup in an asynchronous repository method.
- Examples
- Tests a transfer of 120 between accounts holding 100 each, then checks that both balances remain unchanged.
- What you can do after
- Apply it via EF Core transactions to reverse failed transfers and clear stale tracked state.
Architecting an Apache Iceberg Lakehouse
An operation that makes an earlier table snapshot the current state without necessarily deleting files introduced by later snapshots.
- How it is used
- Supports recovery from bad writes, accidental deletions, dropped partitions, unintended schema changes, and data corruption.
- How deep it goes
- Explains immutable snapshots, metadata-pointer movement, time-travel contrast, and catalog coordination, but not a complete cleanup procedure.
- Examples
- Uses snapshot IDs or timestamps for historical queries and the rollback_to_snapshot procedure for recovery.
- What you can do after
- Diagnose incidents with historical snapshots and restore a previous table state for subsequent queries and writes.
Read how Architecting an Apache Iceberg Lakehouse explains rollback →
Think Like a Software Engineering Manager
A rollback is the restoration of a software system or deployment to a previous known state after a change causes problems.
- How it is used
- Restores production stability after a deployment produces unwanted customer effects.
- How deep it goes
- Defines a prepared response and gives one database-remediation example, without detailing broader implementation mechanics.
- Examples
- Uses duplicate customer records and database queries as the concrete restoration example.
- What you can do after
- Operate/configure it by preparing and executing a predefined procedure that restores a stable production state.
Read how Think Like a Software Engineering Manager explains rollback →
Kubernetes in Action, Second Edition
A rollback is the reversion of a deployed application workload to an earlier version or revision.
- How it is used
- Reverts failed Deployment or StatefulSet updates to an earlier revision while following, or exposing, the workload’s update strategy.
- How deep it goes
- Explains commands, revision inspection, rollout timing, paused Deployments, and manual OnDelete Pod replacement, but not revision storage internals.
- Examples
- Uses kubectl rollout undo, rollout history, targeted revision commands, and manually deleted StatefulSet Pods as operational examples.
- What you can do after
- Operate/configure rollback with kubectl, inspect revisions, select an older Deployment revision, and manually complete OnDelete StatefulSet rollback.
Read how Kubernetes in Action, Second Edition explains rollback →
Acing the Certified Kubernetes Administrator Exam, Second Edition
Rollback is the restoration of an application to a previous known configuration or version after a change.
- How it is used
- Supports rapid recovery when a Kubernetes Deployment change introduces an application bug.
- How deep it goes
- Explains revision history, annotations, undo behavior, and status verification, but does not detail rollback internals.
- Examples
- Uses an Apache Deployment, image-tag change, rollout history, change-cause annotations, undo output, and rollout-status verification.
- What you can do after
- Operate/configure Deployments with kubectl rollout undo, annotate rollback reasons, and verify successful recovery.
Read how Acing the Certified Kubernetes Administrator Exam, Second Edition explains rollback →
GitHub Actions in Action: Continuous integration and delivery for DevOps
Rollback is the process of returning a deployment or system to a previously known good version or state.
- How it is used
- Uses health, metric, and release controls to stop or reverse an unsafe deployment.
- How deep it goes
- Explains rollback triggers and operational techniques, but stops short of detailing rollback storage or recovery implementation.
- Examples
- Demonstrates readiness polling, health monitoring, metric comparison, gradual traffic exposure, and feature-toggle reversal.
- What you can do after
- Operate/configure deployment gates and release controls that restore the last known good behavior when signals indicate failure.
Infrastructure as Code, Patterns and Practices: With examples in Python and Terraform
Rollback is the restoration of infrastructure configuration or system state to an earlier known-good version.
- How it is used
- Uses forward-moving IaC changes to restore functionality after failed infrastructure updates.
- How deep it goes
- Explains commit inspection, git revert, and pipeline application, but stops short of general troubleshooting internals.
- Examples
- Uses a removed database role, Git history, git revert, and static-analysis recovery as concrete examples.
- What you can do after
- Operate IaC recovery by reverting commits and rolling forward a known-working configuration through the deployment pipeline.
Related concepts
- roll-forwardcontrasts with · Roll-forward is a recovery technique that corrects a faulty or failed change by applying a subsequent corrective change rather than reverting to a previous state.
- database transactionpart of · A database transaction is a logical unit of database operations whose changes are committed together or rolled back as a whole.
- feature flaguses · A feature flag is a configuration control that conditionally enables, disables, or selects a software feature or behavior.
- load balanceruses · A load balancer is a hardware or software component that distributes incoming network traffic or requests among multiple destinations, such as servers or service instances.
- kubectl rollout undoimplemented by · kubectl rollout undo is a Kubernetes command that reverts a workload to a previous recorded revision.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.