Question 1
Difficulty: medium
Can you walk me through how you would assess an on-premises application portfolio before starting a cloud migration project?
Sample answer
I’d start with a portfolio assessment that looks beyond the technical inventory. First, I’d group applications by business criticality, user demand, dependencies, compliance requirements, and current pain points. Then I’d identify the hosting model, architecture, data sensitivity, integration points, and performance baselines for each app. I also like to talk with application owners and infrastructure teams early, because the real risks often show up in how systems are used, not just how they’re documented. From there, I’d classify each workload using a migration strategy such as rehost, replatform, refactor, retire, or retain. I’d also estimate effort, risk, and business value so stakeholders can make informed decisions. My goal is to create a migration roadmap that balances quick wins with the more complex workloads that need extra planning. That approach keeps the program realistic, aligned to business priorities, and much easier to govern throughout execution.
Question 2
Difficulty: medium
Tell me about a time you had to manage a cloud migration with tight downtime constraints. How did you handle it?
Sample answer
In a previous migration, we had a customer-facing application that could only tolerate a very short outage window because it supported order processing. I worked with the application, database, and network teams to build a detailed cutover plan with dependencies mapped out hour by hour. We ran multiple rehearsal migrations in a lower environment to uncover issues with DNS propagation, certificate handling, and data synchronization timing. I also made sure we had a rollback plan that was simple enough to execute under pressure, because that is where many migrations fail. On the day of cutover, we froze nonessential changes, monitored the migration in real time, and kept business stakeholders updated at each checkpoint. We finished within the allowed window and the application stabilized quickly afterward. What I learned is that tight downtime is manageable when you invest in rehearsal, communication, and a disciplined cutover checklist instead of relying on assumptions.
Question 3
Difficulty: medium
How do you decide whether a workload should be rehosted, refactored, replatformed, or retired during a cloud migration?
Sample answer
I decide based on business value, technical debt, urgency, and the actual fit of the application for cloud services. If the application is stable, time-sensitive, and needs to move quickly, rehosting can be the best first step. If there is a clear opportunity to improve database performance, scalability, or operations without changing the core application logic too much, replatforming is often the right balance. Refactoring makes sense when the application is strategically important and the existing architecture is holding the business back, but I only recommend it when the value justifies the cost and time. Retirement is just as important because many portfolios contain applications that no longer have a business purpose or have been replaced informally. I try to be practical rather than idealistic. The right migration path is the one that reduces risk, supports business goals, and fits the organization’s timeline and operating maturity.
Question 4
Difficulty: hard
What steps do you take to ensure security and compliance are maintained during a cloud migration?
Sample answer
Security and compliance have to be built into the migration plan from the beginning, not added at the end. I start by identifying the data classification, regulatory requirements, and any control gaps between the current environment and the target cloud setup. Then I work with security teams to define identity and access controls, encryption requirements, logging standards, and network segmentation before migration begins. I also pay attention to shared responsibility, because some teams assume the cloud provider handles more than it actually does. During migration, I verify that data transfers are encrypted, privileged access is controlled, and temporary access is removed when no longer needed. After cutover, I confirm that monitoring, alerting, backup, and vulnerability management are operating as expected. For regulated environments, I make sure evidence is captured throughout the process so audits are easier later. My approach is to reduce risk by designing compliance into each phase rather than trying to prove it after the fact.
Question 5
Difficulty: medium
Describe a time when a migration did not go as planned. What did you do, and what did you learn?
Sample answer
I was involved in a migration where an application came up successfully in the cloud, but the response time was far worse than expected once users started testing it. Instead of pushing ahead, I paused the rollout and brought together the network, application, and database teams to isolate the issue. We found that latency between the application tier and a backend dependency was much higher than in the old environment, and the architecture had not been adjusted for that. We corrected the placement of a few services, tuned connection settings, and rechecked the routing path. Because we had a rollback decision point defined in advance, we were able to stabilize the environment without turning the issue into a major incident. The biggest lesson for me was that a successful migration is not just about getting workloads moved. It’s about validating real user experience, watching for hidden dependencies, and being willing to stop and correct course quickly when something is off.
Question 6
Difficulty: hard
How do you approach dependency mapping for a complex migration involving multiple applications and shared services?
Sample answer
Dependency mapping is one of the most important parts of a migration because hidden connections create most of the surprises. I begin by gathering what documentation exists, but I never rely on it alone. I combine interviews with application owners, infrastructure reviews, network flow data, and, when available, application performance monitoring to identify both direct and indirect dependencies. I pay special attention to shared databases, authentication services, file shares, APIs, batch jobs, and external integrations, because these often affect cutover sequencing. Once I have the data, I create a dependency map that shows what must move together, what can move independently, and what needs a temporary bridge during transition. I also validate the map with the business, because they can often confirm which dependencies are critical versus historical. A good map does more than support planning; it helps prevent outages, reduces guesswork, and gives stakeholders confidence that the migration sequence is grounded in reality.
Question 7
Difficulty: medium
How would you handle a situation where business stakeholders want to migrate everything quickly, but your assessment shows the applications are not ready?
Sample answer
I’d handle that by being direct but constructive. I understand the pressure to move quickly, but I would explain the specific risks in terms the business cares about: downtime, data loss, compliance exposure, performance issues, and likely rework if we rush. Then I’d present a phased plan that creates momentum without compromising stability. For example, we could start with low-risk workloads, standardize landing zone controls, and use those early migrations to improve the process before tackling critical systems. I’d also show a realistic schedule with milestones, so stakeholders can see progress instead of hearing only what cannot be done. If there are business deadlines driving the request, I would explore whether some workloads can be moved using a lower-risk approach while others remain on-premises temporarily. In my experience, stakeholders usually respond well when you offer a path forward rather than just saying no. The key is to translate technical concerns into business impact and keep the conversation solution-oriented.
Question 8
Difficulty: hard
What is your process for planning a cloud landing zone before large-scale migration begins?
Sample answer
My landing zone planning starts with the basics that every migrated workload will depend on: identity, networking, logging, governance, security, and account or subscription structure. I want the foundation to be standardized enough to scale but flexible enough to support different application needs. I typically define how access will be managed, how network segmentation will work, what the tagging and naming standards are, and how central logging and alerting will feed into operations. I also plan for guardrails such as policy enforcement, budget controls, and approved service catalogs so teams can move quickly without creating sprawl. If the organization has multiple environments or business units, I make sure the structure supports that from day one. I see the landing zone as an enabling platform, not just an infrastructure exercise. When it’s done well, migration teams spend less time improvising and more time moving workloads safely and repeatably.
Question 9
Difficulty: medium
How do you measure whether a migration was successful after go-live?
Sample answer
I measure success across technical, operational, and business outcomes. Technically, I check whether the migrated workload meets its baseline for availability, response time, error rates, and recovery objectives. Operationally, I look at whether monitoring, incident response, patching, backup, and access management are working in the new environment without adding unnecessary manual effort. On the business side, I want to know whether the migration met its timeline, stayed within budget, and avoided disruption to users or revenue-generating processes. I also pay attention to whether the migration delivered any of the expected benefits, such as improved scalability, reduced infrastructure overhead, or faster release cycles. A migration can be technically complete and still not be successful if support teams are struggling afterward or if the business gained no value. I usually like to hold a post-migration review to capture lessons learned, close out risks, and feed improvements into the next wave of work. That keeps the program continuously improving instead of repeating the same mistakes.
Question 10
Difficulty: hard
What would you do if you discovered a critical application depends on an outdated system that the business forgot to mention during migration planning?
Sample answer
If I discovered an undocumented critical dependency, I would treat it as a planning risk that needs immediate attention, not as an exception to work around casually. First, I’d confirm the dependency and understand exactly how the application uses it, whether it’s a direct integration, a scheduled batch process, a file exchange, or something more subtle like authentication or reporting. Then I’d assess the impact on the migration timeline, the cutover plan, and any security or compatibility concerns. I’d bring the business owner and technical leads together quickly so everyone understands the options. Depending on the situation, we might need to delay the migration, build a temporary integration bridge, or migrate the dependent system as part of the same wave. I’d also update the dependency register and communication plan so the issue is visible going forward. These surprises happen in real migrations, and the best response is calm, structured, and transparent. The main objective is to protect stability while keeping the program moving with clear decisions.
