Rollback Communication Exercises

1 / 5

An incident commander announces on the call: "We're rolling back to v2.14.3."

What does initiating a rollback communicate to the engineering team?

A rollback decision means the team has assessed that reverting to the last known-good version is the fastest path to restoring service. It is different from a roll-forward (hotfix) where a patch is deployed on top of the broken version.

Term	Meaning
roll back	Revert to a previous known-good version of the application
roll forward	Fix the issue by deploying a new corrective version (hotfix)
known-good version	A previously deployed version that is confirmed to be stable
rollback decision gate	The criteria used for deciding whether to roll back or roll forward (severity, known fix, DB migration risk)

2 / 5

During an incident, the on-call lead says: "The team decided to roll forward instead of rolling back."

When is roll-forward preferred over rollback?

Roll-forward is preferred when: the root cause is known, a fix is small and already prepared, and a rollback would be risky because it would also revert a database migration that cannot be undone safely.

Prefer rollback when	Prefer roll-forward when
Root cause unknown or complex	Root cause is identified and fix is small
Fix will take hours	Fix is ready to deploy now
No DB migration in this release	Rollback would revert an irreversible DB migration
Rollback is safe and fast	Feature flag can isolate the issue immediately

Key vocabulary: hotfix deployment, forward fix, patch release, migration rollback risk.

3 / 5

During a rollback assessment, the on-call engineer says: "We need to assess blast radius before reverting."

What does "blast radius" mean in this rollback context?

Blast radius = the scope of potential impact. Before deciding to rollback, assess: how many users are affected; what downstream services depend on this version; and whether the rollback itself carries a blast radius (does reverting break something that was previously fixed?).

Term	Meaning
blast radius assessment	Estimating scope of impact before taking action — who/what is affected
impact scope	The breadth of affected users, services, or regions
downstream dependencies	Services that depend on the affected component and may be impacted by the rollback
rollback risk vs. staying risk	Comparing the risk of reverting against the risk of remaining on the broken version

4 / 5

An on-call engineer raises a hotfix PR and says it needs an expedited review.

Which statement correctly describes a "hotfix release" compared to a "standard release"?

A hotfix follows an accelerated path: minimum QA scope (targeted regression), expedited approval (synchronous review rather than async), targeted change (smallest possible diff), and a dedicated tracking ticket. The pipeline is shortened, not skipped.

Concept	Standard release	Hotfix release
QA cycle	Full regression suite	Targeted regression only
Review	Async, multiple reviewers, standard SLA	Synchronous, expedited, compensating approval
Change scope	Multiple features / fixes	Single targeted critical fix

Key vocabulary: emergency deployment, expedited review, hotfix branch, change freeze exception.

5 / 5

A change management ticket states: "The deployment is being backed out due to elevated error rates."

What does "back out" mean in deployment vocabulary?

"Back out" is an ITIL/ITSM term for rolling back a change. A backout plan must be documented before a change is approved. It describes specifically how to revert the change if something goes wrong.

Term	Meaning
backout / back out	Revert a change to the previous state — synonymous with rollback, used in ITIL contexts
backout plan	A pre-documented, step-by-step procedure for reverting a change, required before CAB approval
backout window	The time period after deployment during which the backout can be safely executed
deployment window	The approved time slot for making the change, including the backout window