Request-based auto scaling in GCP - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ service_behavior

intermediate

How does request-based auto scaling respond to traffic spikes?

In Google Cloud Run, when request-based auto scaling is enabled, what happens if your service suddenly receives a large number of requests?

A Cloud Run rejects new requests until the traffic decreases to the previous level.

B Cloud Run instantly creates new instances to handle all incoming requests without delay.

C Cloud Run queues all requests until the existing instances finish processing, then creates new instances.

D Cloud Run gradually adds new instances based on the number of concurrent requests per instance until the traffic is handled efficiently.

💡 Hint

Think about how Cloud Run balances load and scales instances smoothly.

Cloud Run uses request-based auto scaling to add instances gradually based on concurrent requests per instance. It does not instantly create all instances nor reject requests.

❓ Configuration

intermediate

Configuring max instances for request-based auto scaling

You want to limit the maximum number of instances your Cloud Run service can scale to during high traffic. Which configuration setting controls this limit?

A Setting the <code>min-instances</code> flag to zero in the Cloud Run service configuration.

B Setting the <code>max-instances</code> flag in the Cloud Run service deployment configuration.

C Setting the <code>concurrency</code> value to 1 in the Cloud Run service configuration.

D Setting the <code>timeout</code> value to the lowest possible number.

💡 Hint

Think about which setting limits the upper bound of instances.

The max-instances setting limits how many instances Cloud Run can create, controlling maximum scale.

❓ Architecture

advanced

Designing a Cloud Run service for efficient request-based auto scaling

You have a Cloud Run service that processes image uploads. To optimize cost and performance, which architecture choice best supports efficient request-based auto scaling?

A Set concurrency to 1 and allow max-instances to scale to 100, so each instance handles one request at a time but many instances can run.

B Set concurrency to 1 and max-instances to 1, so only one instance handles all requests sequentially.

C Set concurrency to a higher number like 80 and limit max-instances to a moderate number like 10, so each instance handles multiple requests concurrently.

D Set concurrency to 1000 and max-instances to 1, so one instance handles all requests concurrently.

💡 Hint

Think about balancing instance count and concurrency for cost and performance.

Higher concurrency with moderate max-instances allows efficient use of resources and smooth scaling.

❓ security

advanced

Security considerations with request-based auto scaling in Cloud Run

When using request-based auto scaling in Cloud Run, what is a key security best practice to protect your service from sudden traffic spikes caused by malicious requests?

A Use Cloud Armor or similar tools to filter and block malicious traffic before it reaches Cloud Run.

B Increase the request timeout to allow long-running requests to complete.

C Set max-instances to unlimited to handle any traffic volume without restrictions.

D Disable authentication to allow all requests to pass through quickly for scaling.

💡 Hint

Think about how to prevent bad traffic from causing resource waste.

Using Cloud Armor helps filter malicious traffic, protecting your service and controlling scaling costs.

✅ Best Practice

expert

Optimizing request-based auto scaling for cost and latency

You want to optimize your Cloud Run service to balance cost and latency during unpredictable traffic. Which combination of settings and practices achieves this best?

A Set concurrency moderate, max-instances moderate, and enable minimum instances to reduce cold starts while controlling cost.

B Set concurrency high, max-instances low, and disable minimum instances to save cost but accept cold starts.

C Set concurrency to 1, max-instances high, and enable minimum instances to keep some instances always warm.

D Set concurrency low, max-instances low, and disable minimum instances to minimize resource usage.

💡 Hint

Think about trade-offs between keeping instances warm and controlling max scale.

Moderate concurrency and max-instances with minimum instances enabled balances latency and cost effectively.