IT Operations, Risk and Audit · #100
Online IT Change and Rollback Plan
Schedules steps, reserves rollback time and produces the plan for the CAB as Markdown.
- Data is processed in your browser and not sent to a server
- Export: MD
Tool
What does this tool do?
The most common mistake in maintenance windows is continuing implementation until the last minute without reserving rollback time. The tool schedules implementation steps by duration, subtracts rollback time from the window to calculate the point of no return and creates a plan with checklists.
How to use it
- Enter title, systems, start, window length, risk and approvers.
- Enter implementation and rollback steps as
minutes;descriptionone per line; add the rollback trigger, validation checks and communication steps. - Press Run; download the Markdown plan and attach it to the change record.
Fields
- Change title
- Affected systems
- Start
- Local time.
- Window (min)
- Risk
- Approvers
- Implementation steps
- minutes;description — one step per line.
- Rollback steps
- minutes;description
- Rollback trigger
- Under what condition to roll back?
- Validation checks
- One check per line.
- Communication
- One step per line.
Worked example
| Input | Value |
|---|---|
| Change title | Exchange 2019 CU15 update |
| Affected systems | EX01, EX02 (DAG) |
| Start | 2026-11-28T22:00 |
| Window (min) | 150 |
| Risk | Medium |
| Approvers | CAB |
| Implementation steps | |
| Rollback steps | |
| Rollback trigger | Mail flow test fails or services do not start within 15 minutes |
| Validation checks | |
| Communication | |
Expected result: 5 implementation steps take 100 min, rollback 25 min; 25 min of slack in a 150-minute window. Point of no return 00:05: if validation fails at this time, rollback must start.
The plan downloads as Markdown with timeline, checklist, rollback steps and communication and can be pasted into Teams, a wiki or the ticket system. Times are calculated from the start time; use the Maintenance Window tool for time zone differences.
Second scenario
| Input | Value |
|---|---|
| Change title | Core switch firmware |
| Affected systems | SW-CORE-01 |
| Start | 2026-12-05T01:00 |
| Window (min) | 60 |
| Risk | High |
| Implementation steps | |
| Rollback steps | |
Expected result: Core switch firmware: 50 min implementation + 20 min rollback do not fit a 60-minute window (10 min short); a high-risk change also has no approver.
Method and assumptions
Step start = sum of previous steps. Point of no return = start + (window − rollback time). Slack = window − implementation − rollback; negative gives a warning.
Using the result and troubleshooting
Share the plan before the CAB meeting and record real durations for the next change. Check the SLA budget of affected services with the SLA tool.
Limits and common mistakes
- Steps are sequential; parallel steps and dependencies are not modelled. The plan content is user input; technical correctness is not checked.
Frequently asked questions
Why does the point of no return matter?
If rollback starts after this time it cannot finish within the window and an unplanned outage results. Plan validation to finish before this point.
Related tools and service
- Maintenance Window Time Zone PlannerConverts the maintenance time to local time for each location, flags business-hour overlaps and creates an invite file.
- IT Risk MatrixScores information security risks, places them on a heat map and flags missing ownership or treatment.
Related service: Change, Configuration and Patch Management Processes