A rollback is a recovery operation that restores data or system state to a previously valid state by discarding or reversing changes.
Rollback is a recovery response when a production deployment fails to start or causes unacceptable performance. Deployability scenarios use it to restore service without violating SLAs. section 5.5.1
Rollback is triggered by detected problems such as:
A configuration problem that prevents the updated component from starting. section 5.5.1
Unacceptable REST endpoint latency after deployment, attributed to an optimized query. section 5.5.1
How Microservices Patterns, Second Edition covers it
How deep it goes
Shows two failure-triggered rollback scenarios but does not specify rollback mechanics.
Examples
Uses configuration-related startup failure and unacceptable REST endpoint latency after query optimization.
What it assumes
Assumes a tested deployment pipeline and detected failures, without teaching those mechanisms.
What you can do after
Evaluate rollback requirements for deployment failures while keeping SLAs intact.
How it connects
Connects rollback to deployment pipelines, production deployment, and service-level agreements.
Architecting an Apache Iceberg LakehouseAn operation that makes an earlier table snapshot the current state without necessarily deleting files introduced by later snapshots.
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.
transactional fixtureused by · A test fixture whose database changes are enclosed in a transaction that is rolled back during teardown to restore the starting state.
feature flaguses · A feature flag is a configuration control that conditionally enables, disables, or selects software functionality at build time or runtime.
database transactionpart of · A database transaction is a bounded unit of database operations whose changes are treated as one logical whole, typically committed or rolled back together.
deployment pipelineused by · A deployment pipeline is a repeatable sequence of stages that builds, tests or validates, packages, and delivers software changes toward release or deployment.
This explainer was generated by Manning's knowledge graph from the text of Microservices Patterns, Second Edition. The section links point to the parts of the book it draws on.