GCP ACE Udemy 1 Flashcards
Your app-building company has several projects under its umbrella. Different teams at your enterprise have different cloud Budgets and their GCP billing is managed by different billing accounts. Your company has a centralized finance team that needs a single visual representation of all costs incurred. New cost data should be included in the reports as soon as it’s available. What should you do?
A. Export the billing data to BigQuery using Billing data export and create a Data Studio dashboard for visualization
B. Export the costs to CSVC from Costs table and visualize it using Data Studio
C. Use the pricing calculator to get the pricing on a per resource basis
D. Go to the Cloud Billing Console Reports to view the desire costs information
A. Export the billing data to BigQuery using Billing data export and create a Data Studio dashboard for visualization
Explanation:
A is correct because you can run an analysis on Bigquery after exporting the billing reports from all projects to the same dataset. It suggests using Billing data export to export the billing data to BigQuery and then create a Data Studio dashboard for visualization. This option allows for the centralization of all costs incurred from different teams and billing accounts, providing a single visual representation for the finance team.
B is incorrect because CSV export is a manual process and not very efficient and scalable. It suggests exporting the costs to CSV from the Costs table. This option may not provide real-time data updates and would require manual efforts to export and visualize the data in Data Studio. It does not fulfill the requirement of including new costs data as soon as it’s available.
C is incorrect because we need actual prices and not just estimates. It suggests using the pricing calculator to get pricing on a per-resource basis. The pricing calculator is useful for estimating costs but not for visualizing the actual costs incurred across different projects and billing accounts. It does not fulfill the requirement of providing a single visual representation for the finance team.
D is incorrect because the reports view will not show billing information of all projects in the same window as the billing accounts are different. It suggests going to the Cloud Billing Console Reports view. While the Cloud Billing Console provides some reporting capabilities, it may not provide the flexibility and customization options required to create a comprehensive visual representation of all costs incurred. It does not fulfill the requirement of creating a single visual representation for the finance team.
You work at a billion-dollar RPG game development company. You are running your Gaming server on GKE on multiple pods running on four n1-standard-2 nodes on a GKE cluster. Additional pods need to be deployed on the same cluster requiring an n2-highmem-16 type of node. Your app is live in production and cannot afford downtime. What should you do?
A. Run gcloud container clusters upgrade before deploying the new services
B. 1. Create a new Code Pool with n2-highmen-16 machine type. 2. Deploy the new pods
C. 1. Create a new cluster with n2-highmen-16 nodes. 2. Delete the old cluster and redploy the pods in the new cluster
D. 1. Create a new cluster with both n1-standard-2 and n2-highmen-16 nodes. 2. Delete the old cluster and redeploy the pods
B. 1. Create a new Code Pool with n2-highmen-16 machine type. 2. Deploy the new pods
Explanation:
A is incorrect because you need to create a new node pool for the new pods as they require different types of instances. Running “gcloud container clusters upgrade” would upgrade the existing nodes in the cluster to a different machine type. This would not meet the requirement of deploying new pods on nodes with the n2-highmem-16 machine type.
B is correct because you can add new types of instances to the GKE cluster by adding node pools. It will not cause any downtime to the existing cluster. Creating a new Node Pool with the n2-highmem-16 machine type allows for the deployment of new pods on the same cluster. This option does not involve any downtime as the existing pods can continue running on the current nodes while the new pods are deployed on the new Node Pool.
C is incorrect because there is no need to create a new cluster for it. Creating a new cluster with n2-highmem-16 nodes and deleting the old cluster would result in downtime during the redeployment of the pods. The app being live in production cannot afford downtime.
D is incorrect because there is no need to create a new cluster for it. Creating a new cluster with both n1-standard-2 and n2-highmem-16 nodes and deleting the old cluster would also result in downtime during the redeployment of the pods. Additionally, it would require managing and maintaining two different node types in the cluster, which may not be necessary or efficient.
You work as a data scientist in an e-commerce shoe-selling company. Your website uses Cloud Spanner as its database backend to keep current state information about users. All events triggered by the users are logged in Cloud Bigtable. The Cloud Spanner data is exported every day to Cloud Storage for backup purposes. Your Datascience team is training an ML model on the user data and they need to join data from Cloud Spanner and Bigtable together. How can you fulfill this requirement as efficiently as possible?
A. Copy data from Cloud Storage and Cloud Bigtable for specific users using a dataflow job
B. Create a dataflow job that copies data from Cloud Bigtable and Cloud Spanner for specific users
C. Write a Spark job that runs on a Dataproc cluster to extract data from Cloud Bigtable and Cloud Storage for specific users
D. 1. Create two separate BigQuery external tables on Cloud Storage and Cloud Bigtable 2. Join these tables through user gields using the BigQuery console and apply appropriate filters
D. 1. Create two separate BigQuery external tables on Cloud Storage and Cloud Bigtable 2. Join these tables through user fields using the BigQuery console and apply appropriate filters
Explanation:
A is incorrect because creating a dataflow job can require significant effort. It suggests copying data from Cloud Storage and Cloud Bigtable for specific users using a dataflow job. This approach involves unnecessary data movement and processing, as it requires copying data from two separate sources and then filtering it based on specific users.
B is incorrect because creating a dataflow job can require significant effort. It suggests creating a dataflow job that copies data from Cloud Bigtable and Cloud Spanner for specific users. Similar to option A, this approach involves unnecessary data movement and processing, as it requires copying data from two separate sources and then filtering it based on specific users.
C is incorrect because using Dataproc and Spark can require significant effort and time. It suggests writing a Spark job that runs on a Dataproc cluster to extract data from Cloud Bigtable and Cloud Storage for specific users. This approach also involves unnecessary data movement and processing, as it requires extracting data from two separate sources and then filtering it based on specific users.
D is correct because Bigquery supports analytics on data through external tables from Cloud Storage and Bigtable. It is perfect for this use case. It suggests creating two separate BigQuery external tables on Cloud Storage and Cloud Bigtable. By doing so, the data from both sources can be directly queried and joined in BigQuery without the need for additional data movement or processing. The join can be performed using the user fields, and appropriate filters can be applied to retrieve the required data efficiently. This approach minimizes data movement, reduces processing overhead, and provides an efficient way to join data from both Cloud Spanner and Cloud Bigtable.
