Launching a virtual machine in the cloud

WHAT THIS TESTS This checks baseline cloud literacy. The interviewer wants to confirm you know the knobs that define a VM and that you instinctively think about security and reproducibility.

A GOOD ANSWER COVERS Walk the configuration in order. Start with the machine image, which fixes the operating system and any preinstalled software. Pick an instance type that maps to a CPU, memory, and network profile suited to the workload. Place the instance in a VPC and subnet, deciding whether it needs a public IP, and attach a security group that allows only the ports you intend. Attach block storage, choosing the volume size and performance class, and decide whether the root volume persists on termination. Finally, set up access through an SSH key pair or a cloud identity role, and ideally bake the setup into an image or IaC template so it is repeatable.

COMMON WRONG ANSWERS Forgetting the security group and exposing SSH or RDP to the entire internet. Ignoring whether the subnet is public or private. Embedding long-lived credentials on the box instead of using an instance role. Treating storage as an afterthought.

LIKELY FOLLOW-UPS How do you connect to a private instance. What is the difference between instance store and persistent block storage. How would you make this launch reproducible. How do you grant the VM permission to call other cloud services.

ONE CONCRETE EXAMPLE To run a small web app you select a current Linux image, a general-purpose instance type, and place it in a public subnet with a security group allowing only HTTPS from anywhere and SSH from your office IP. You attach a modest SSD volume, assign an instance role granting read access to a config bucket, and launch with your key pair, capturing the whole setup in a launch template for next time.