AWS CloudFormation Support
Recovering a CloudFormation stack in UPDATE_ROLLBACK_FAILED.
A failed update normally rolls back by itself. UPDATE_ROLLBACK_FAILED means the rollback also hit a resource it could not restore, and the stack is stuck until a person fixes the cause. We find the blocking resource, repair it and continue the rollback.
When You Need This
The situations where this work pays for itself.
Signs it is time
- The stack shows UPDATE_ROLLBACK_FAILED and will not accept updates
- A stack is stuck in ROLLBACK_FAILED after a failed create
- A resource was changed or deleted outside CloudFormation
- A deployment keeps failing and rolling back
- A rollback failed because of permissions or a dependency
What you get
- A root-cause note naming the resource and failure reason
- The stack returned to a healthy, updatable state
- A change set based deployment process
- Guardrails against manual changes to managed resources
How we do it
A clear, safe process you can follow along with.
1. Read the stack events
We open the Events and find the first resource with a failed status and its status reason. That reason, not the stack-level state, tells us what blocked the rollback.
2. Understand why rollback failed
Typical causes are a resource modified or deleted manually, an IAM role that no longer allows the action, a dependency such as a security group still in use, or a resource limit reached.
3. Repair the cause in place
We restore what rollback expected: recreate the deleted resource with matching properties, fix the IAM permission, detach the dependency or free the quota, depending on the reason.
4. Continue the update rollback
With the cause fixed, we run continue-update-rollback in the console or CLI. The stack returns to UPDATE_ROLLBACK_COMPLETE if every resource can be restored.
5. Skip resources only as a last resort
If a resource cannot be restored, we can continue the rollback while skipping it, which marks it as updated or failed in the template state. We note the drift this creates and clean up manually afterward.
6. Fix the original failure
We find why the update failed, test the change with a change set or in a test stack, and redeploy so the same problem does not recur.
Mistakes we help you avoid
Common ways this goes wrong, and what we do instead.
Skipping resources immediately
It leaves the stack and real resources out of sync. Repair first.
Fixing things by hand without recording
Manual repairs create drift unless we update the template.
Deleting the stack to escape
It may delete data. Verify retention policies first.
Updating without a change set
Change sets preview replacements and deletions before they hit production.
Frequently Asked Questions
What does UPDATE_ROLLBACK_FAILED mean in CloudFormation?
An update failed and CloudFormation could not roll back to the previous state, usually because a resource cannot be restored. The stack stays blocked until you fix it and continue the rollback.
How do I fix a stack stuck in UPDATE_ROLLBACK_FAILED?
Read the events to find the failed resource, correct the cause, then run continue update rollback. Skipping resources is a last option that may cause drift.
How do I stop updates from failing and rolling back?
Use change sets, validate templates, test in a lower environment, and avoid editing managed resources outside CloudFormation.
Want Help With Recovering a CloudFormation stack in UPDATE_ROLLBACK_FAILED?
Schedule an appointment. Tell us about your environment and we will outline the safest, highest-value approach.
Schedule Appointment › No commitment required