What is an allow list?
An application control policy where ONLY explicitly approved applications are permitted to run. Everything else is blocked by default. Most restrictive — unknown = blocked. Also called a whitelist.
What is a deny list?
An application control policy that blocks specifically named applications while allowing everything else. Antivirus is a large deny list. More flexible than allow list but less secure — new threats may not be listed yet.
What is scope discipline in technical change management?
Sticking strictly to the approved change scope during implementation. If additional issues are discovered, document them and submit separate change requests. Do not expand scope mid-window without explicit authorization.
What is scope creep?
The unauthorized expansion of a change beyond its approved scope during implementation. Dangerous because the expanded scope hasn't been reviewed, tested, or approved — significantly increasing risk.
What are the 4 restart types in order of impact?
1. Service restart — stop/start one service (fastest)
2. Application restart — close/relaunch one app
3. OS reboot — full operating system restart
4. Power cycle — physical power off/on (most disruptive)
What is the primary/secondary failover strategy for downtime?
Switch users to secondary system → upgrade primary → verify → switch users back. Minimizes downtime for critical services. Often automated. If the primary upgrade fails, simply point users back to secondary (which hasn't been changed).
What is a legacy application in the context of change management?
An older system, often no longer supported by its developer, running for many years. Has undocumented behavior and dependencies. Any nearby changes risk breaking it. Must be carefully documented before touching the surrounding environment.
What is a dependency?
A requirement where one system/service needs another to function. Example: upgrading firewall management software may require firewalls to run a newer firmware first. Dependencies can span multiple systems and must be identified before the change window.
Why use batched deployment?
Deploy to a small group first (e.g., 50 of 400 systems). Verify success. Then proceed to the remainder. Limits blast radius — if the first batch fails, only a small number of systems are affected before the issue is caught.
What is version control in change management?
Tracking changes to configuration files, scripts, and settings over time. Lets admins see what changed between any two points and revert to a known-good state if a change causes problems. Essential for configurations — not just software code.
Why must documentation be updated after every change?
Outdated documentation is as dangerous as no documentation. Future staff will trust what's written and make decisions based on it. If it's wrong, they'll make the wrong decisions. Documentation should reflect current state — always.
An approved change is for printer driver upgrades on workstations. During the window, the technician discovers the print server also needs an update. What should happen?
Complete the approved workstation driver upgrade. Document the print server issue. Submit a separate change request for the print server update. Do NOT expand scope mid-window without authorization.
What should you do before modifying a system with a legacy application?
Study and document the application and how it's installed. Understand its dependencies. Clone the environment in a sandbox and test your change there. Document findings so the team is no longer afraid of it — you become the expert.
What does antivirus have in common with a deny list?
Antivirus is effectively a large deny list — it allows everything to run EXCEPT code that matches known malicious signatures. Everything not specifically identified as bad is allowed. This is the fundamental difference from an allow list.
Why does a service restart help validate change resilience?
Performing a planned service restart during a maintenance window reveals how a system behaves after a stop/start — which is the same behavior as a service crash recovery or power outage. It's a controlled test of system recovery.
What types of items should be in version control?
Router/switch/firewall configurations · Registry changes (Windows) · GPO templates · Application configuration files · Scripts · OS patch state · Any configuration that changes over time and needs to be auditable or reversible.