Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A global retail organization is migrating its virtualized workloads from an on-premises data center to Compute Engine. Currently, routine hardware maintenance and host operating system patching in their on-premises environment cause significant operational challenges:
How does Compute Engine's live migration technology address these challenges during planned infrastructure maintenance?
Cross-regional load balancing is a capability of Cloud Load Balancing that distributes incoming user requests across backend virtual machine (VM) instances deployed in multiple geographic regions. Powered by Google's global network and edge Points of Presence (PoPs), global external HTTP(S) load balancing routes traffic to the healthy region closest to the user and automatically diverts traffic to an alternate region if an entire region becomes unreachable.
Cross-regional traffic diversion is a high availability (HA) and disaster recovery (DR) architecture pattern designed to handle large-scale regional outages or traffic spikes, not the mechanism live migration uses during routine single-host maintenance. Live migration keeps the existing VM running by moving it to another host server within the same zone, without rebooting the VM or requiring standby instances in another region.
The scenario asked how an organization can architect a globally distributed application using Cloud Load Balancing to remain accessible under a single external IP address if an entire Google Cloud region experiences an outage.
While combining regional managed instance groups (MIGs) and global load balancing is a best practice for resilient system design, live migration operates automatically at the Infrastructure as a Service (IaaS) host layer within a single zone, even for standalone VMs without a load balancer.
A managed instance group (MIG) is a collection of identical virtual machine (VM) instances managed as a single entity using an instance template. Its autohealing feature continuously monitors the health of each VM by running application-based health checks. If an application crashes, freezes, or stops responding to health check probes, the MIG automatically deletes the unhealthy VM and recreates a brand-new instance to replace it.
Live migration does not delete and recreate virtual machines when a physical host server undergoes planned patching. Recreating a VM destroys the existing instance's in-memory state and requires booting a fresh instance from an image. Instead, live migration moves the existing, running VM to another physical host in the same zone before the underlying server goes offline, preserving the active instance without replacement.
The scenario asked which Compute Engine feature automatically detects when an application process inside a VM has crashed or become unresponsive and replaces the faulty instance with a healthy copy.
Students often confuse autohealing with live migration because both features improve workload availability automatically. However, autohealing recovers from application or guest OS failures by recreating VMs, whereas live migration avoids planned host infrastructure downtime by moving running VMs intact.
Live migration is a built-in Compute Engine capability that automatically moves running virtual machine (VM) instances away from physical host servers scheduled to undergo infrastructure maintenance. Its primary purpose is to keep virtualized workloads operational and make routine data center maintenance—such as hardware repairs, power grid updates, and host operating system (OS) or BIOS patching—completely transparent to customers.
When a physical server requires maintenance or exhibits signs of impending hardware failure, Google Cloud's cluster management software transfers the complete application state and guest memory of the running VM to a healthy target host within the same zone. This addresses the retail organization's challenges by eliminating scheduled downtime windows, removing the need for manual IT intervention, and preserving all network configurations without rebooting the guest operating system.
metadata, block storage data, and network settings remain completely unchanged.Unlike traditional maintenance strategies that rely on terminating instances, rebooting servers, or failing over entire application stacks to secondary regions, live migration handles host-level events transparently at the infrastructure layer. It enables high availability for individual VMs by default, allowing businesses to maintain continuous operations 24/7/365 without complex failover orchestration.
Stopping an instance and restarting it on another physical server describes the terminate and restart (TERMINATE and RESTART) host maintenance policy in Compute Engine, rather than live migration. When a VM is configured with this policy, Google Cloud halts the running instance when a host maintenance event begins and boots it back up on a healthy host server that is not involved in the maintenance activity.
This approach does not keep virtual machines running continuously without rebooting during routine host events. Stopping and restarting a VM causes an interruption to active processes, clears volatile in-memory application state, and results in brief workload downtime while the guest OS shuts down and reboots on the new host.
RDMA-enabled compute-optimized VMs—do not support live migration and must use TERMINATE and RESTART.The scenario asked how Compute Engine handles host maintenance for specialized workloads that do not support live migration (such as GPU-accelerated VMs), or for applications explicitly configured to avoid temporary migration performance dips by rebooting on a fresh host.
A common misconception is that all cloud maintenance requires a VM reboot. While Compute Engine offers TERMINATE as a configurable host maintenance option for fault-tolerant clusters that demand constant maximum performance, live migration is the default behavior for standard VMs specifically to avoid reboots.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated