Computing in the Cloud CLOUD-DIGITAL-LEADER Practice Questions (Page 4)

Questions 16–20

1. 16application · medium

   A software company is refactoring a monolithic application into smaller, independently deployable components. Each component will be containerized and needs to scale independently based on its own traffic patterns. The company also needs to manage service discovery and load balancing between these components. Which compute approach best meets these requirements?

   • AUse a single serverless function for all components and rely on the platform to route requests internally.
   • BDeploy each component as a separate VM and use a network load balancer to distribute traffic.
   • CContainerize each component and deploy them on a Kubernetes cluster with an ingress controller for load balancing.
   • DRun all components in a single container and use a load balancer to distribute traffic to that container.
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   What is load balancing in cloud computing?

   • AA technique that automatically scales resources based on demand
   • BA technique that reduces the number of compute resources to save costs
   • CA technique that stores data across multiple storage devices
   • DA technique that distributes incoming traffic across multiple compute resources to improve performance and reliability
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   An e-commerce website experiences variable traffic, with high demand during promotional events and lower demand at other times. The company wants to ensure that the website remains responsive during traffic spikes without over-provisioning resources during normal periods. What is the best approach to manage compute resources?

   • ADeploy the website on a fixed number of large VMs to handle peak traffic at all times.
   • BManually add more VMs during promotional events and remove them afterward.
   • CUse a single powerful VM with a load balancer to distribute traffic to different application instances on the same machine.
   • DUse a managed instance group with an autoscaling policy based on CPU utilization or request count.
   Reveal Solution Discussion Select an answer first
5. 19expert · hard

   A media company runs a video transcoding pipeline. The pipeline consists of several stages: some stages are long-running (up to 30 minutes) and require GPU acceleration, while others are short and event-driven. The company wants to minimize operational overhead and cost, but also needs to handle unpredictable spikes in demand. They are considering a mix of compute options. Which combination is most appropriate?

   • AUse preemptible VMs with GPUs for the long-running transcoding stages and serverless functions for the short, event-driven stages.
   • BUse serverless functions for all stages, including the GPU-accelerated ones.
   • CUse a Kubernetes cluster with GPU nodes for all stages, including the short ones.
   • DUse standard VMs with GPUs for all stages and rely on autoscaling to handle demand.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A data analytics company has developed a Python application that uses a specific version of a library and requires a particular operating system configuration. They want to deploy this application to multiple cloud environments and on-premises servers without rewriting the code or changing the environment setup. What is the best way to achieve this portability?

   • APackage the application and its dependencies into a container image and run it on any container runtime.
   • BRewrite the application to be cloud-agnostic and use a serverless function in each environment.
   • CCreate a custom VM image with the application and dependencies pre-installed and deploy it to each environment.
   • DUse a configuration management tool to install the dependencies on each server manually.
   Reveal Solution Discussion Select an answer first