High availability configuration in GCP - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Architecture

intermediate

Designing a multi-region high availability setup

You want to design a high availability architecture for a web application on Google Cloud Platform. Which setup ensures the application remains available even if an entire region fails?

A Deploy the application in multiple zones within a single region without load balancing.

B Deploy the application in a single region with multiple zones and use a regional load balancer.

C Deploy the application in multiple regions with global load balancing and health checks.

D Deploy the application in a single zone with autoscaling and use a zonal load balancer.

💡 Hint

Think about what happens if a whole region goes down. Which option covers multiple regions?

Only deploying across multiple regions with global load balancing ensures availability if one region fails. Regional or zonal setups protect against smaller failures but not entire region outages.

❓ Configuration

intermediate

Configuring Cloud SQL for high availability

You want to configure a Cloud SQL instance for high availability. Which configuration option correctly enables automatic failover?

A Create a primary instance with a failover replica in a different zone and enable high availability.

B Create multiple read replicas in the same zone without failover enabled.

C Create a single primary instance without replicas and enable backups.

D Create a primary instance and manually switch to a backup instance when needed.

💡 Hint

Automatic failover requires a failover replica in a different zone and enabling HA.

Cloud SQL high availability requires a failover replica in a different zone and enabling the HA option to allow automatic failover in case the primary fails.

❓ security

advanced

Securing a multi-region high availability setup

In a multi-region high availability setup, which practice best protects your application from unauthorized access while maintaining availability?

A Use a single firewall rule allowing all traffic to all regions.

B Disable all firewalls and rely on network peering for security.

C Use a VPN connection only to one region and allow open access elsewhere.

D Use Identity-Aware Proxy (IAP) with global load balancers and enforce least privilege IAM roles.

💡 Hint

Think about controlling access securely and limiting permissions.

Using IAP with global load balancers and least privilege IAM roles ensures secure access control without compromising availability across regions.

❓ service_behavior

advanced

Behavior of global load balancer during regional failure

What happens to user traffic when a Google Cloud global HTTP(S) load balancer detects that the primary region is down?

A Traffic is automatically routed to healthy backend instances in other regions without user impact.

B Traffic continues to be sent to the primary region until manually rerouted.

C Traffic is dropped until the primary region recovers.

D Traffic is routed randomly to any backend regardless of health status.

💡 Hint

Global load balancers monitor backend health and reroute traffic automatically.

The global HTTP(S) load balancer automatically routes traffic to healthy backends in other regions when it detects a regional failure, ensuring continuous availability.

✅ Best Practice

expert

Optimizing database replication for high availability and latency

You have a multi-region application with a Cloud Spanner database. To optimize for both high availability and low latency, which configuration is best?

A Use a single regional instance with synchronous replication to a failover region.

B Use a multi-region instance with replicas distributed across regions and configure read-write transactions to the closest region.

C Use multiple independent regional instances without replication and synchronize data manually.

D Use a single regional instance with asynchronous replication to multiple regions.

💡 Hint

Consider how Cloud Spanner handles replication and latency across regions.

Cloud Spanner multi-region instances distribute replicas across regions, providing high availability and allowing read-write transactions to be served with low latency by routing to the closest replica.