Chapter 1 GCP ACE Wiley Flashcards
Which of the following is an option for choosing a computing resource in Google Cloud?
A. Cache
B. Virtual machine (VM)
C. Block
D. Subnet
B. Virtual machine (VM)
Explanation:
A basic unit for purchasing computing resource sin Google Cloud is the virtual machine. Cache is incorrect because this is a low latency storage system. Block is incorrect because a block is a unit of storage on persistent disks. A subnet is incorrect; a subnet is network abstraction
If you use a cluster that is managed by a cloud provider, which of these will be managed for you by the cloud provider?
A. Monitoring
B. Networking
C. Some security management tasks
D. All of the above
D. All of the above
Explanation:
When using managed clusters, the cloud provider will monitor the health of servers, also known as nodes, in the cluster; set up networking between nodes in the cluster and configure firewalls and other security controls
You need serverless computing for file processing and running the back end of a website, which two products can you choose from Google Cloud?
A. Kubernetes Engine and Compute Engine
B. Cloud Run and Cloud Functions
C. Cloud Functions and Compute Engine
D. Cloud Functions and Kubernetes Engine
B. Cloud Run and Cloud Functions
Explanation:
Cloud Run is a serverless platform for running containers, and Cloud Functions is a service for executing short running functions in response to events. Kubernetes Engine is a managed cluster service, and both Kubernetes Engine and Compute Engine require you to configure servers. Neither Compute Engine nor Kubernetes is a serverless option
You have been asked to design a storage system for a web application that allows users to upload large data files to be analyzed by a data analytics workflow. The files should be stored in a high availability storage system. Filesystem functionality is not required. Which storage system in Google Cloud should be used?
A. Block Storage
B. Object Storage
C. Cache
D. Network File System
B. Object Storage
Explanation:
Object storage, like Cloud Storage, provides redundantly stored objects without limits on the amount of data you cans tore, which makes option B correct. Since file system functionality is not required, option D is not a good option. Block storage could be used, but you would have to manage your own replication to ensure high availability and it would cost more than object storage. Caches are transient, in memory storage and are not high availability, persistent storage systems
All block storage systems use what block size?
A. 4 KB
B. 8 KB
C. 16 KB
D. Block size can vary
D. Block size can vary
Explanation:
The correct answer is D. Block sizes in blocks troage system can vary. Block size is established when a filesystem is created. In Linux, 4KB block sizes are commonly used
You have been asked to set up network security in a virtual private cloud. Your company wants to have multiple subnetworks and limit traffic between the subnetworks. Which network security control would you use to control the flow of traffic between subnets?
A. Identity access management
B. Router
C. Firewall
D. IP Address Table
C. Firewall
Explanation:
Firewalls in Google Cloud are software defined network controls that limit the flow of traffic into and out of a network or subnetwork.
Routers are used to move traffic to appropriate destinations on the network. Identity Access Management is used for authenticating and authorizing users; it is not relevant to network controls between subnets. IP address tables are not a security control
When you create a machine learning service to learn how to classify objects using tabular data, what type of servers should you choose to manage compute resources?
A. Virtual machines
B. Clusters of VMs
C. No servers; you should use specialized services, which are serverless
D. VMs running Linux only
C. No servers; you should use specialized services, which are serverless
Explanation:
Specialized services in Google Cloud, like AutoML are serverless. Google manages the compute resources used by the services. There is no need for a user to allocate or manage servers
When doe sinvesting in servers for extended periods of time, such as committing to use servers for three to five years work well?
A. When a company is just starting up
B. When a company can accurately predict server need for an extended period of time
C. When a company has a fixed IT budget
D. When a company has a variable IT budget
B. When a company can accurately predict server need for an extended period of time
Explanation:
Investing in servers works well when an organization can accurately predict the number of servers and other equipment it will need for an extended period of time and can utilize that equipment consistently. Startups are not established businesses with histories that can guide expected needs in three to five years. It does not matter if a budget is fixed or variable; investing in servers should be based on demand for sever capacity
Your company is based in North America and will be running a virtual server for batch processing invoices. What factor determines the unit per minute cost?
A. The time of day the virtual machine is run
B.. The characteristics of the server
C. The application you run
D. None of the above
B.. The characteristics of the server
Explanation:
The characteristics of the server, such as the number of virtual servers, the amount of memory and the regfion where you run the VM, influence the cost, so option B is correct. Time of day is not a factor, nor is the type of application you run on the VM
You plan to use AutoML to analyze sales data and predict product demand in the near future. You plan to analyze between 1,000 and 2,500 products per hour. How many VMs should you allocate to meet peak demand?
A. 1
B. 10
C. 25
D. None; AutoML is a serverless service
D. None; AutoML is a serverless service
Explanation:
AutoML is one of Google Clouds specialized services. Users of the service do not need to configure any VMs to use the service
You have to run a number of services to support an application. Which of the following is a good deployment model?
A. Run on a large, single VM
B. Use containers in a managed cluster
C. Use two large VMs, making one of them read only
D. Use a small VM for all services and increase the size of the VM when CPU utilization exceeds 90 percent
B. Use containers in a managed cluster
Explanation:
Containers give the most flexibility for using the resources of a cluster efficiently, and orchestration platforms reduce the operations overhead, which makes option B correct. Running in a single VM is not recommended because if the server fails, all services will be down. Using two VMs with one read only is not useful. Read only is sometimes used with databases, but there was no mention of databases in question. Using a small VM and upgrading when it is no longer able to keep up with the workload delivers poor quality service to users and should be avoided
You have created a VM. Which of the following system administrator operations are you allowed to perform on it?
A. Configure the file system
B. Patch operating system software
C. Change file and directory permissions
D. All of the above
D. All of the above
Explanation:
All of the operations are available to a system administrator after creating a VM
Cloud Filestore is based on what filesystem technology?
A. Network File System (NFS)
B. XFS
C. EXT4
D. ReiserFS
A. Network File System (NFS)
Explanation:
Cloud Filestore is based on the Network File System (NFS), which is a distributed file management system. The other options are filesystems supported by Linux
When creating resources in Google Cloud, those resources are always part of what?
A. Virtual Private cloud
B. Subdomain
C. Cluster
D. None of the above
A. Virtual Private cloud
Explanation:
When you create resources, they are created within a VPC. Resources are added to the VPC and are not accessible outside the VPC unless you explicitly configure them to be. A subdomain is related to web domains and not related to the organization of Google Cloud resources. Clusters, such as Kubernetes clusters may be in your network, but not all resources are necessarily in a cluster
You need to store data for an application and are using a cache. How will the cache affect data retrieval?
A. A cache improves the execution of client side JavaScript
B. A cache will continue to store data even if power is lost, improving availability
C. Caches can get out of sync with the system of truth
D. Using a cache will reduce latency, since retrieving from a cache is faster than retrieving from SSDs or HDDs
D. Using a cache will reduce latency, since retrieving from a cache is faster than retrieving from SSDs or HDDs
Explanation: Caches use memory and that makes them the fastest storage type for reading data. Caches are data stores on the back end of distributed systems, not the clients. A cache would have no effect on client side JavaScript execution. Caches can lose data in the cache if power is lost and the data would have to be reloaded. Caches can get out of sync with the system of truth because the system of truth could be updated, but the cache may not be updated. Caches have faster read times than SSDs and HDDs