You need to create the network infrastructure to deploy a highly available web application in the us-east1 and us-west1 regions. The application runs on Compute Engine instances, and it does not require the use of a database. You want to follow Google-recommended practices. What should you do?
To deploy a highly available web application across multiple regions, following Google-recommended practices, you should create one VPC with one subnet in each region and use a global load balancer with a static IP address. The global load balancer ensures that traffic can be efficiently managed across regions, providing high availability. Enabling Cloud CDN and Google Cloud Armor on the load balancer optimizes content delivery and enhances security. Finally, creating an A record using the IP address of the load balancer in Cloud DNS consolidates the setup, ensuring ease of access to the web application.
100% B. https://cloud.google.com/solutions/best-practices-compute-engine-region-selection#distributed_frontend_and_backend_in_multiple_regions
100% B.
B as you need Global Load Balancer across multiple Regions.
D would have been correct if it did not have the Standard network tiering. Remember, standard network tier: Network services such as Cloud Load Balancing are regional (one VIP per region) - https://cloud.google.com/network-tiers C doesn't make sense - You don't need peering with GSLB. This leaves 1 and 2. You need GSLB for CDN. Hence that eliminates A.
Its B not C
B is 100% correct based on given scenario.
• B. Create one VPC with one subnet in each region. Create a global load balancer ***** with a static IP address. Enable Cloud CDN and Google Cloud Armor on the load balancer. Create an A record using the IP address of the load balancer in Cloud DNS.
option D is the correct answer as it suggests creating one VPC with one subnet in each region and using a global HTTP(S) load balancer with Cloud CDN enabled. This follows the Google-recommended practice of having a VPC in each region for higher availability and lower latency, while also providing global load balancing and CDN acceleration for the web application.
Option A is incorrect because it suggests using a regional network load balancer in each region, which would not provide global load balancing across both regions. Also, Cloud CDN cannot be enabled on regional load balancers, only on global load balancers. Option B is incorrect because it suggests using a global load balancer, which would not require a VPC in each region. However, this does not follow the recommended practice of having a VPC in each region for higher availability and lower latency. Also, Cloud CDN and Google Cloud Armor can only be enabled on global HTTP(S) load balancers, not on network load balancers.
Option C is incorrect because it suggests peering two VPCs in each region, which is not necessary for deploying a highly available web application in multiple regions. Also, Cloud CDN can only be enabled on global HTTP(S) load balancers, not on network load balancers. Lastly, creating a CNAME record for the load balancer is not recommended as it adds extra DNS resolution steps and can negatively impact performance.
Where did you read that Google suggest to have one VPC for each region for HA? Anyways, also in D you have just 1 VPC with 2 regional subnet
Same Opinion for me: https://cloud.google.com/load-balancing/docs/load-balancing-overview?hl=en "Cloud CDN is supported with the global external HTTP(S) load balancer and the global external HTTP(S) load balancer (classic)." So we need a HTTP/S Loadbalancer
Wrong - cannot use CDN with Standard tier
There's NO SUCH THING as a CNAME pointing to a IP addresses, CNAMEs must ALWAYS point to a FQDN. D is wrong.