HackMD
GCP ACE Problems 002
Hint
- The image file may be corrupted
- The server hosting the image is unavailable
- The image path is incorrect
- The image format is not supported
- Your company's infrastructure is on-premises, but all machines are running at maximum capacity. You want to burst to Google Cloud. The workloads on Google
Cloud must be able to directly communicate to the workloads on-premises using a private IP range. What should you do?
A. In Google Cloud, configure the VPC as a host for Shared VPC.
B. In Google Cloud, configure the VPC for VPC Network Peering.
C. Create bastion hosts both in your on-premises environment and on Google Cloud. Configure both as proxy servers using their public IP addresses.
D. Set up Cloud VPN between the infrastructure on-premises and Google Cloud.
最佳解是: D
建立Cloud VPN是最佳选择，因为它将在Google Cloud和公司基础架构之间建立一个安全的、加密的连接。这样，公司基础架构上的工作负载和Google Cloud上的工作负载就可以使用私有IP范围直接通信。Cloud VPN还提供了灵活的拓扑，可以满足各种需要。此外，Cloud VPN不需要在Google Cloud中进行大量的配置或更改。选项D是最佳选择。
选项A不是最佳选择，因为Shared VPC是一个托管VPC子网的服务，它允许共享VPC中的资源和服务。但是，Shared VPC并不提供直接连接到公司基础架构的功能。
选项B不是最佳选择，因为VPC网络对等连接只能在同一区域的VPC之间建立连接。而且，VPC对等连接不能直接连接到公司基础架构，因为这需要在公共Internet上使用公共IP地址。
选项C不是最佳选择，因为在公司基础架构和Google Cloud中都需要设置堡垒主机，并且使用它们的公共IP地址作为代理服务器。这不仅增加了复杂性，还降低了安全性，因为公共IP地址容易受到攻击和黑客攻击。此外，使用代理服务器会增加延迟和网络带宽的消耗。
- You want to select and configure a solution for storing and archiving data on Google Cloud Platform. You need to support compliance objectives for data from one geographic location. This data is archived after 30 days and needs to be accessed annually. What should you do?
A. Select Multi-Regional Storage. Add a bucket lifecycle rule that archives data after 30 days to Coldline Storage.
B. Select Multi-Regional Storage. Add a bucket lifecycle rule that archives data after 30 days to Nearline Storage.
C. Select Regional Storage. Add a bucket lifecycle rule that archives data after 30 days to Nearline Storage.
D. Select Regional Storage. Add a bucket lifecycle rule that archives data after 30 days to Coldline Storage.
最佳解是: D
选项D是最佳选择，因为这需要满足合规性目标，根据问题描述，这需要满足特定地理位置的数据归档需求。Regional Storage存储桶只能在同一地理区域内使用，因此可以满足地理位置的要求。对于归档数据，Coldline Storage是最经济实惠的选择，可以帮助降低存储成本。因此，添加一个bucket生命周期规则，以便在30天后归档数据到Coldline Storage，是最佳选择。
选项A不是最佳选择，因为Multi-Regional Storage是一种跨地理位置的存储桶类型。它适合用于存储需要高可用性和低延迟的数据。但是，这并不满足题目中特定地理位置的要求。
选项B不是最佳选择，因为Nearline Storage是一种低成本的对象存储解决方案，适合用于频繁访问但不需要长时间保存的数据。因为问题描述中指定了数据需要保存一年，因此这不是最佳选择。
选项C不是最佳选择，因为Regional Storage和Nearline Storage都不满足问题描述中特定地理位置的要求。此外，Nearline Storage虽然是一种低成本的解决方案，但对于归档数据不是最优解。
- Your company uses BigQuery for data warehousing. Over time, many different business units in your company have created 1000+ datasets across hundreds of projects. Your CIO wants you to examine all datasets to find tables that contain an employee_ssn column. You want to minimize effort in performing this task.
What should you do?
A. Go to Data Catalog and search for employee_ssn in the search box.
B. Write a shell script that uses the bq command line tool to loop through all the projects in your organization.
C. Write a script that loops through all the projects in your organization and runs a query on INFORMATION_SCHEMA.COLUMNS view to find the employee_ssn column.
D. Write a Cloud Dataflow job that loops through all the projects in your organization and runs a query on INFORMATION_SCHEMA.COLUMNS view to find employee_ssn column.
最佳解是: A
在Data Catalog中搜索employee_ssn是最简单的方法，可以快速找到包含该列的所有表格和数据集。其他选项需要编写脚本或编写代码来循环遍历所有项目，并且需要一些编程技能和时间。
- You create a Deployment with 2 replicas in a Google Kubernetes Engine cluster that has a single preemptible node pool. After a few minutes, you use kubectl to examine the status of your Pod and observe that one of them is still in Pending status:
What is the most likely cause?
A. The pending Pod's resource requests are too large to fit on a single node of the cluster.
B. Too many Pods are already running in the cluster, and there are not enough resources left to schedule the pending Pod.
C. The node pool is configured with a service account that does not have permission to pull the container image used by the pending Pod.
D. The pending Pod was originally scheduled on a node that has been preempted between the creation of the Deployment and your verification of the Pods' status. It is currently being rescheduled on a new node.
最佳解是: D
如果 Pod 的狀態為 Pending，表示該 Pod 尚未分配到任何 Node。在這個情境中，最有可能的原因是 D 選項中提到的問題。由於你使用的是單個折扣節點池，該節點可能已經在 Pod 分配後被折扣回收。當出現這種情況時，該 Pod 的狀態將為 Pending，直到該 Pod 被重新調度到另一個 Node。其他選項提到的問題通常不會導致 Pod 保持在 Pending 狀態。例如，即使 Pod 的資源需求超出節點的可用資源，Kubernetes 仍會調度該 Pod，直到有足夠的資源可用。
- You want to find out when users were added to Cloud Spanner Identity Access Management (IAM) roles on your Google Cloud Platform (GCP) project. What should you do in the GCP Console?
A. Open the Cloud Spanner console to review configurations.
B. Open the IAM & admin console to review IAM policies for Cloud Spanner roles.
C. Go to the Stackdriver Monitoring console and review information for Cloud Spanner.
D. Go to the Stackdriver Logging console, review admin activity logs, and filter them for Cloud Spanner IAM roles.
最佳解是: D
在GCP控制台中，通过Stackdriver Logging控制台中的管理活动日志可以查看Cloud Spanner IAM角色的信息。 Stackdriver Logging记录所有GCP活动，这包括Cloud Spanner项目的管理员活动日志。在Stackdriver Logging中，您可以设置过滤器以筛选Cloud Spanner项目的管理员活动。
其他选项：
A. Cloud Spanner控制台可用于查看配置，但无法提供有关IAM角色分配的详细信息。
B. IAM和管理控制台可用于查看IAM策略，但是不提供有关IAM角色分配的详细信息。
C. Stackdriver监视可用于监视基础结构和应用程序，并提供与Cloud Spanner相关的信息，但是不能提供IAM角色分配的详细信息。
- Your company implemented BigQuery as an enterprise data warehouse. Users from multiple business units run queries on this data warehouse. However, you notice that query costs for BigQuery are very high, and you need to control costs. Which two methods should you use? (Choose two.)
A. Split the users from business units to multiple projects.
B. Apply a user- or project-level custom query quota for BigQuery data warehouse.
C. Create separate copies of your BigQuery data warehouse for each business unit.
D. Split your BigQuery data warehouse into multiple data warehouses for each business unit.
E. Change your BigQuery query model from on-demand to flat rate. Apply the appropriate number of slots to each Project.
最佳解是: B和E
解釋：
A. 將商務部門的使用者分散到多個專案中可能會使管理和維護變得困難。對成本沒有影響。
B. 通過應用用戶或專案級別的自定義查詢配額，可以控制用戶的查詢花費，進而控制成本。
C. 在每個業務部門中創建BigQuery數據倉庫的副本會占用大量存儲空間，也難以維護。
D. 這樣做也會導致管理和維護變得困難，同時也會導致額外的成本。
E. 通過將BigQuery查詢模型從按需轉換為平坦費用模型，並為每個專案應用適當的slot數量，可以更好地控制成本。
GCP Official Examples
- Your organization plans to migrate its financial transaction monitoring application to Google Cloud. Auditors need to view the data and run reports in BigQuery, but they are not allowed to perform transactions in the application. You are leading the migration and want the simplest solution that will require the least amount of maintenance. What should you do?
A. Assign roles/bigquery.dataViewer to the individual auditors.
B. Create a group for auditors and assign roles/viewer to them.
C. Create a group for auditors, and assign roles/bigquery.dataViewer to them.
D. Assign a custom role to each auditor that allows view-only access to BigQuery.
最佳解: C
A 不正確，因為 Google 推薦的做法是將 IAM 角色分配給組，而不是個人。組比單個用戶更容易管理，並且它們提供了對角色和權限的高級別可見性。
B 不正確，因為它使用基本角色來授予審核員對項目所有資源的查看權限。
C 是正確的，因為它使用預定義角色為審計員組提供對 BigQuery 的查看訪問權限。如果工作職責發生變化，可以在組中添加或刪除審核員。
D 不正確，因為使用預定義的角色可以實現目標並且需要較少的維護。
- You are managing your company’s first Google Cloud project. Project leads, developers, and internal testers will participate in the project, which includes sensitive information. You need to ensure that only specific members of the development team have access to sensitive information. You want to assign the appropriate Identity and Access Management (IAM) roles that also require the least amount of maintenance. What should you do?
A. Assign a basic role to each user.
B. Create groups. Assign a basic role to each group, and then assign users to groups.
C. Create groups. Assign a Custom role to each group, including those who should have access to sensitive data. Assign users to groups.
D. Create groups. Assign an IAM Predefined role to each group as required, including those who should have access to sensitive data. Assign users to groups.
最佳解: D
A 不正確有兩個原因：推薦的做法是使用組而不是為每個用戶分配角色。除此之外，基本角色沒有足夠的粒度來說明對敏感數據的訪問。
B 不正確，因為基本角色沒有足夠的粒度來說明對敏感數據的訪問。
C 不正確，因為創建和維護自定義角色比使用預定義角色需要更多維護。
D 是正確的，因為預定義角色的粒度足夠細，可以為需要訪問敏感數據的特定角色設置權限。此解決方案還使用組，這是管理單個角色權限的推薦做法。
- You are responsible for monitoring all changes in your Cloud Storage and Firestore instances. For each change, you need to invoke an action that will verify the compliance of the change in near real time. You want to accomplish this with minimal setup. What should you do?
A. Use the trigger mechanism in each datastore to invoke the security script.
B. Use Cloud Function events, and call the security script from the Cloud Function triggers.
D. Use a Python script to get logs of the datastores, analyze them, and invoke the security script.
C. Redirect your data-changing queries to an App Engine application, and call the security script from the application.
最佳解: B
A 不正確，因為在每個單獨的數據庫中設置觸發器需要額外的設置。
B 是正確的，因為它提供快速響應並且需要最少的設置。
C 是不正確的，因為它需要自定義編程。
D 不正確，因為它需要大量的自定義編程。
Your application needs to process a significant rate of transactions. The rate of transactions exceeds the processing capabilities of a single virtual machine (VM). You want to spread transactions across multiple servers in real time and in the most cost-effective manner. What should you do?
A. Send transactions to BigQuery. On the VMs, poll for transactions that do not have the ‘processed’ key, and mark them ‘processed’ when done.
B. Set up Cloud SQL with a memory cache for speed. On your multiple servers, poll for transactions that do not have the ‘processed’ key, and mark them ‘processed’ when done.
C. Send transactions to Pub/Sub. Process them in VMs in a managed instance group.
D. Record transactions in Cloud Bigtable, and poll for new transactions from the VMs.
最佳解: C
A 是不正確的，因為它的延遲明顯高於所需的實時響應。
B 不正確，因為它不會提供所需的性能。
C 是正確的，因為 Pub/Sub 是一種可擴展的解決方案，可以以低成本在多個服務器之間有效地分發大量任務。
D 是不正確的，因為雖然速度很快，但它會引入額外的存儲數據開銷。
Your team needs to directly connect your on-premises resources to several virtual machines inside a virtual private cloud (VPC). You want to provide your team with fast and secure access to the VMs with minimal maintenance and cost. What should you do?
A. Set up Cloud Interconnect.
B. Use Cloud VPN to create a bridge between the VPC and your network.
C. Assign a public IP address to each VM, and assign a strong password to each one.
D. Start a Compute Engine VM, install a software router, and create a direct tunnel to each VM.
最佳解: B
A 是不正確的，因為它比其他現有解決方案貴得多。
B 是正確的，因為它符合 Google 推薦的做法。
C 是不正確的，因為它需要相當大的維護工作。
D 不正確，因為為每個單獨的 VM 設置連接需要大量維護。
You are implementing Cloud Storage for your organization. You need to follow your organization’s regulations. They include: 1) Archive data older than one year. 2) Delete data older than 5 years. 3) Use standard storage for all other data. You want to implement these guidelines automatically and in the simplest manner available. What should you do?
A. Set up Object Lifecycle management policies.
B. Run a script daily. Copy data that is older than one year to an archival bucket, and delete five-year-old data.
C. Run a script daily. Set storage class to ARCHIVE for data that is older than one year, and delete five-year-old data.
D. Set up default storage class for three buckets named: STANDARD, ARCHIVE, DELETED. Use a script to move the data in the appropriate bucket when its condition matches your company guidelines.
最佳解: A
A 是正確的，因為對像生命週期允許您自動執行組織的數據策略。
B 不正確，因為更改對象的存儲類別不需要將對象複製到另一個存儲桶。
C 是不正確的，因為它需要自定義編程。
D 不正確，因為將對象移動到 DELETED 存儲桶並沒有真正刪除它。
You are creating a Cloud IOT application requiring data storage of up to 10 petabytes (PB). The application must support high-speed reads and writes of small pieces of data, but your data schema is simple. You want to use the most economical solution for data storage. What should you do?
A. Store the data in Cloud Spanner, and add an in-memory cache for speed.
B. Store the data in Cloud Storage, and distribute the data through Cloud CDN for speed.
C. Store the data in Cloud Bigtable, and implement the business logic in the programming language of your choice.
D. Use BigQuery, and implement the business logic in SQL.
最佳解: C
A 不正確，因為 Cloud Spanner 不是最經濟的解決方案。
B 不正確，因為面向 blob 的 Cloud Storage (雲存儲)不適合讀取和寫入小塊數據。
C 是正確的，因為 Bigtable 提供高速讀寫，容納簡單的模式，並且具有成本效益。
D 不正確，因為 BigQuery 不提供物聯網所需的高速讀寫。
You have created a Kubernetes deployment on Google Kubernetes Engine (GKE) that has a backend service. You also have pods that run the frontend service. You want to ensure that there is no interruption in communication between your frontend and backend service pods if they are moved or restarted. What should you do?
A. Create a service that groups your pods in the backend service, and tell your frontend pods to communicate through that service.
B. Create a DNS entry with a fixed IP address that the frontend service can use to reach the backend service.
C. Assign static internal IP addresses that the frontend service can use to reach the backend pods.
D. Assign static external IP addresses that the frontend service can use to reach the backend pods.
最佳解: A
A 是正確的，因為 Kubernetes 服務的目的是提供一個目的地，當 pod 移動或重新啟動時可以使用該目的地。
B 不正確，因為 DNS 條目是由服務創建創建的。
C 不正確，因為當 pod 重新啟動時，靜態內部 IP 地址不會自動更改。
D 不正確，因為當 Pod 重新啟動時，靜態外部 IP 地址不會自動更改，並且它們會將流量帶到 Google 網絡之外。
You are responsible for the user-management service for your global company. The service will add, update, delete, and list addresses. Each of these operations is implemented by a Docker container microservice. The processing load can vary from low to very high. You want to deploy the service on Google Cloud for scalability and minimal administration. What should you do?
A. Deploy your Docker containers into Cloud Run.
B. Start each Docker container as a managed instance group.
C. Deploy your Docker containers into Google Kubernetes Engine.
D. Combine the four microservices into one Docker image, and deploy it to the App Engine instance.
最佳解: A
A 是正確的，因為 Cloud Run 是一項需要最少管理的託管服務。
B 不正確，因為託管實例組缺乏公開其服務的管理能力。
C 是不正確的，因為儘管 GKE 提供了可擴展性，但它需要對集群進行持續管理。
D 不正確，因為它需要努力在一個 Docker 容器中重新實現四個微服務。你也會失去你的微服務架構。
You provide a service that you need to open to everyone in your partner network. You have a server and an IP address where the application is located. You do not want to have to change the IP address on your DNS server if your server crashes or is replaced. You also want to avoid downtime and deliver a solution for minimal cost and setup. What should you do?
A. Create a script that updates the IP address for the domain when the server crashes or is replaced.
B. Reserve a static internal IP address, and assign it using Cloud DNS.
C. Reserve a static external IP address, and assign it using Cloud DNS.
D. Use the Bring Your Own IP (BYOIP) method to use your own IP address.
最佳解: C
A 不正確，因為更新 DNS 記錄最多可能需要 24 小時，並且會導致停機。
B 不正確，因為內部 IP 不可路由且無法在 Internet 上看到。
C 是正確的，因為外部 IP 是可路由的，可以在互聯網上發布和看到，這也是最具成本效益的解決方案。
D 不正確，因為儘管有可能，但使用您自己的 IP 地址並不像 Google Cloud DNS 那樣具有成本效益。
Your team is building the development, test, and production environments for your project deployment in Google Cloud. You need to efficiently deploy and manage these environments and ensure that they are consistent. You want to follow Google-recommended practices. What should you do?
A. Create a Cloud Shell script that uses gcloud commands to deploy the environments.
B. Create one Terraform configuration for all environments. Parameterize the differences between environments.
C. For each environment, create a Terraform configuration. Use them for repeated deployment. Reconcile the templates periodically.
D. Use the Cloud Foundation Toolkit to create one deployment template that will work for all environments, and deploy with Terraform.
最佳解: D
A 不正確，因為創建符合 Google Cloud 推薦做法的 gcloud 命令的自定義腳本需要大量的開發和維護工作。
B 不正確，因為參數化環境差異既費時又容易出錯。
C 是不正確的，因為它容易出錯並且涉及大量的對賬工作。
D 是正確的，因為 Cloud Foundation Toolkit (CFT) 提供了反映 Google Cloud 推薦做法的現成模板，可用於自動創建環境。
You receive an error message when you try to start a new VM: “You have exhausted the IP range in your subnet.” You want to resolve the error with the least amount of effort. What should you do?
A. Create a new subnet and start your VM there.
B. Expand the CIDR range in your subnet, and restart the VM that issued the error.
C. Create another subnet, and move several existing VMs into the new subnet.
D. Restart the VM using exponential backoff until the VM starts successfully.
最佳解: B
A 不正確，因為您不需要新的子網。擴展 CIDR 範圍後，初始 VM 將通過重新部署來工作。
B 是正確的，因為一旦您擴展了 CIDR 範圍，您就可以重新部署它，並且它會起作用。
C 不正確，因為將您的 VM 移動到另一個子網是一項額外的耗時工作，而不是必需的。
D 不正確，因為一旦 CIDR 範圍用盡，重新部署出現故障的 VM 將無法解決問題。
You are running several related applications on Compute Engine virtual machine (VM) instances. You want to follow Google-recommended practices and expose each application through a DNS name. What should you do?
A. Use the Compute Engine internal DNS service to assign DNS names to your VM instances, and make the names known to your users.
B. Assign each VM instance an alias IP address range, and then make the internal DNS names public.
C. Assign Google Cloud routes to your VM instances, assign DNS names to the routes, and make the DNS names public.
D. Use Cloud DNS to translate your domain names into your IP addresses.
最佳解: D
A 不正確，因為電子郵件不是提交 DNS 發布請求的方式。
B 不正確，因為您不能公開內部 DNS 名稱。
C 不正確，因為您不能公開 DNS 名稱。
D 是正確的，因為 Cloud DNS 是將域名轉換為 IP 地址的正確工具。
You are charged with optimizing Google Cloud resource consumption. Specifically, you need to investigate the resource consumption charges and present a summary of your findings. You want to do it in the most efficient way possible. What should you do?
A. Rename resources to reflect the owner and purpose. Write a Python script to analyze resource consumption.
B. Attach labels to resources to reflect the owner and purpose. Export Cloud Billing data into BigQuery, and analyze it with Data Studio.
C. Assign tags to resources to reflect the owner and purpose. Export Cloud Billing data into BigQuery, and analyze it with Data Studio.
D. Create a script to analyze resource usage based on the project to which the resources belong. In this script, use the IAM accounts and services accounts that control given resources.
最佳解: B
A 是不正確的，因為它需要自定義編程並且不遵循 Google 推薦的做法並且不是最有效的解決方案。
B 是正確的，因為它描述了 Google 推薦的做法：標籤附加到資源，然後這些標籤傳播到計費項目中。
C 不正確，因為在為資源創建標籤時不再創建標籤，並且不能用於跟踪資源。
D 不正確，因為它需要自定義編程。
You are creating an environment for researchers to run ad hoc SQL queries. The researchers work with large quantities of data. Although they will use the environment for an hour a day on average, the researchers need access to the functional environment at any time during the day. You need to deliver a cost-effective solution. What should you do?
A. Store the data in Cloud Bigtable, and run SQL queries provided by Bigtable schema.
B. Store the data in BigQuery, and run SQL queries in BigQuery.
C. Create a Dataproc cluster, store the data in HDFS storage, and run SQL queries in Spark.
D. Create a Dataproc cluster, store the data in Cloud Storage, and run SQL queries in Spark.
最佳解: B
A 不正確，因為 HBase 不允許臨時查詢。
B 是正確的，因為 BigQuery 允許臨時查詢並且具有成本效益。
C 不正確，因為 HDFS 不是推薦用於 Google Cloud 上的 Dataproc 的存儲。
D 不正確，因為它不是最具成本效益的解決方案，因為集群一直在運行。
You are migrating your workload from on-premises deployment to Google Kubernetes Engine (GKE). You want to minimize costs and stay within budget. What should you do?
A. Configure Autopilot in GKE to monitor node utilization and eliminate idle nodes.
B. Configure the needed capacity; the sustained use discount will make you stay within budget.
C. Scale individual nodes up and down with the Horizontal Pod Autoscaler.
D. Create several nodes using Compute Engine, add them to a managed instance group, and set the group to scale up and down depending on load.
最佳解: A
A 是正確的，因為 Autopilot 旨在降低管理集群的運營成本並針對生產優化集群。
B 不正確，因為它違反了按需供應而不是過度供應的原則。雖然持續使用折扣會降低預算，但不使用不必要的資源會進一步降低成本。
C 是不正確的，因為 Horizontal Pod Autoscaler 是為了調整 Kubernetes 參數以提高性能，而不是為了取出不必要的資源。
D 不正確，因為儘管 Google Kubernetes Engine 在內部使用 Compute Engine，但託管實例組缺乏用於擴展 Kubernetes 的 Autopilot 功能。
Your application allows users to upload pictures. You need to convert each picture to your internal optimized binary format and store it. You want to use the most efficient, cost-effective solution. What should you do?
A. Store uploaded files in Cloud Bigtable, monitor Bigtable entries, and then run a Cloud Function to convert the files and store them in Bigtable.
B. Store uploaded files in Firestore, monitor Firestore entries, and then run a Cloud Function to convert the files and store them in Firestore.
C. Store uploaded files in Filestore, monitor Filestore entries, and then run a Cloud Function to convert the files and store them in Filestore.
D. Save uploaded files in a Cloud Storage bucket, and monitor the bucket for uploads. Run a Cloud Function to convert the files and to store them in a Cloud Storage bucket.
最佳解: D
A 不正確，因為 BigTable 對存儲二進製文件有限制。
B 不正確，因為 Firestore 對於大型二進製文件效率不高。
C 不正確，因為它不是最具成本效益的解決方案。
D 是正確的，因為它遵循 Google 推薦的做法，並且是最有效、最具成本效益的解決方案。
You are migrating your on-premises solution to Google Cloud. As a first step, the new cloud solution will need to ingest 100 TB of data. Your daily uploads will be within your current bandwidth limit of 100 Mbps. You want to follow Google-recommended practices for the most cost-effective way to implement the migration. What should you do?
A. Set up Partner Interconnect for the duration of the first upload.
B. Obtain a Transfer Appliance, copy the data to it, and ship it to Google.
C. Set up Dedicated Interconnect for the duration of your first upload, and then drop back to regular bandwidth.
D. Divide your data between 100 computers, and upload each data portion to a bucket. Then run a script to merge the uploads together.
最佳解: B
A 不正確，因為合作夥伴互連雖然比專用互連便宜，但仍然不是此遷移的最具成本效益的解決方案。
B 是正確的，因為它遵循 Google 針對這些數據大小推薦的做法，並且是實施遷移的最具成本效益的解決方案。
C 不正確，因為專用互連對於此用例而言不是最具成本效益的。
D 不正確，因為它不是最具成本效益的解決方案。
You are setting up billing for your project. You want to prevent excessive consumption of resources due to an error or malicious attack and prevent billing spikes or surprises. What should you do?
A. Set up budgets and alerts in your project.
B. Set up quotas for the resources that your project will be using.
C. Set up a spending limit on the credit card used in your billing account.
D. Label all resources according to best practices, regularly export the billing reports, and analyze them with BigQuery.
最佳解: B
A 是不正確的，因為預算和警報會導致通知，但不會阻止過度的資源消耗。
B 是正確的，因為設置配額將防止資源消耗超過指定的限制。
C 是不正確的，因為它不會防止過多的資源消耗。相反，您的信用卡將產生未付餘額；您會收到一封來自 Google 的電子郵件，但仍需支付費用。
D 不正確，因為分析超出預算的根本原因並不能防止超支。
Your project team needs to estimate the spending for your Google Cloud project for the next quarter. You know the project requirements. You want to produce your estimate as quickly as possible. What should you do?
A. Build a simple machine learning model that will predict your next month’s spend.
B. Estimate the number of hours of compute time required, and then multiply by the VM per-hour pricing.
C. Use the Google Cloud Pricing Calculator to enter your predicted consumption for all groups of resources.
D. Use the Google Cloud Pricing Calculator to enter your consumption for all groups of resources, and then adjust for volume discounts.
最佳解: C
A 是不正確的，因為雖然 ML 在許多領域產生了很好的結果，但有更直接的方法需要更少的時間來產生估計。
B 不正確，因為您需要添加其他費用，例如存儲和數據出口費用。
C 是正確的，因為 Google Cloud 定價計算器可以快速給出結果，並且您知道該項目所需的資源。
D 不正確，因為批量折扣（也稱為持續使用折扣）會自動應用並包含在計算器估算中。
