Which approach minimizes user impact during fixes when debugging a live system?

• A. Deploying a full fix during peak hours without notification
• B. Using non-blocking techniques, feature flags, and staged rollouts to minimize user impact
• C. Taking the service offline until the fix is complete
• D. Releasing a temporary workaround without testing

Answer: (B)

Whenever debugging a live system, the aim is to fix without disrupting users. Using non-blocking techniques, feature flags, and staged rollouts achieves this because fixes can be applied without halting traffic, can be turned on gradually, and can be rolled back quickly if issues arise. Non-blocking design means ongoing requests aren’t blocked; feature flags let you enable the fix for a subset of users and monitor impact before broader exposure; staged rollouts distribute the change across users to catch problems early with minimal fallout. This approach reduces blast radius, supports rapid monitoring, and allows safe, fast rollback if needed. In contrast, deploying a full fix during peak hours without notification forces users through potential instability; taking the service offline eliminates availability; releasing a temporary workaround without testing is risky and can introduce new issues.