Cloud Run service concept in GCP - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Cloud Run Service Scaling Behavior

What happens when a Cloud Run service receives a sudden spike of 1000 requests at the same time?

A Cloud Run rejects all requests exceeding the first instance's capacity without scaling.

B Cloud Run queues all requests and processes them one by one on a single instance.

C Cloud Run automatically creates multiple instances to handle the requests concurrently, scaling up quickly.

D Cloud Run requires manual intervention to add more instances to handle the load.

💡 Hint

Think about how serverless platforms handle sudden traffic.

Cloud Run automatically scales the number of instances based on incoming requests, allowing it to handle sudden spikes without manual intervention.

❓ Architecture

intermediate

Cloud Run Service and Traffic Splitting

You want to deploy a new version of your Cloud Run service and gradually send 20% of traffic to it while keeping 80% on the current version. How can you achieve this?

A Deploy the new version to a separate Cloud Run service and manually route 20% of requests via DNS.

B Use Cloud Run's traffic splitting feature to assign 20% traffic to the new revision and 80% to the old revision.

C Update the existing service to the new version and rely on Cloud Run to automatically split traffic.

D Create a load balancer in front of Cloud Run and configure it to split traffic by percentage.

💡 Hint

Cloud Run supports gradual rollouts using built-in features.

Cloud Run allows you to split traffic between revisions of the same service by percentage, enabling gradual rollouts without extra infrastructure.

❓ security

advanced

Cloud Run Service Authentication Configuration

You want to restrict access to your Cloud Run service so only authenticated users from your Google Workspace domain can invoke it. Which configuration achieves this?

A Set the Cloud Run service to allow only authenticated invocations and configure IAM to grant the 'Cloud Run Invoker' role to your Google Workspace domain users.

B Make the Cloud Run service public and rely on application-level authentication inside the container.

C Disable authentication on Cloud Run and use a firewall to block unwanted IP addresses.

D Use a VPN to restrict network access to the Cloud Run service.

💡 Hint

Think about how Cloud Run integrates with IAM and Google identities.

Cloud Run supports IAM-based authentication. Granting the 'Cloud Run Invoker' role to your domain users and requiring authentication restricts access properly.

❓ service_behavior

advanced

Cloud Run Service Concurrency Setting Impact

If you set the concurrency of a Cloud Run service to 1, what is the effect on instance scaling and request handling?

A Each instance handles only one request at a time, so Cloud Run may create more instances under load to handle concurrent requests.

B Each instance can handle multiple requests simultaneously, reducing the number of instances needed.

C Cloud Run queues requests internally and processes them sequentially on a single instance regardless of concurrency setting.

D Setting concurrency to 1 disables autoscaling and limits the service to one instance.

💡 Hint

Consider how concurrency affects instance utilization and scaling.

Concurrency set to 1 means each instance processes one request at a time, so Cloud Run scales out more instances to handle multiple requests concurrently.

✅ Best Practice

expert

Optimizing Cloud Run Service Startup Time

Your Cloud Run service experiences high latency on the first request after a period of inactivity. Which approach best reduces this cold start delay?

A Use a scheduled Cloud Function to send requests periodically to keep the service active.

B Increase the CPU and memory allocation to the service to speed up startup.

C Deploy the service to multiple regions to distribute the load.

D Use Cloud Run's minimum instance setting to keep a small number of instances always warm and ready.

💡 Hint

Think about how to keep instances ready without manual pings.

Setting minimum instances keeps some instances running even when idle, reducing cold start latency without extra maintenance.