Rollback
A practical DevOps definition, with an example you can use.
Returning a system to an earlier known version or configuration after an unsuccessful change.
An example you can picture
A team re-deploys the previous image when a release introduces errors. It then verifies recovery with application checks and user-facing indicators.
The distinction that matters
Not every change is reversible. Database migrations, deleted data, and external side effects may need a forward fix or separate recovery plan.
Put it into practice
For a lab release, write a recovery plan before changing the application. Name the previous version and a verification check. Identify any data change that would prevent a simple rollback.
Connect the ideas: deployment and runbook. For the bigger picture, read the related learning guide.
Reference
Read the primary documentation for the underlying concept and further detail.