Your organization operates a global, mission-critical application hosted on Google Cloud using Compute Engine Managed Instance Groups (MIGs) behind an external Application Load Balancer. To improve release velocity and prevent customer-facing incidents, the SRE team requires the deployment pipeline to automate canary deployments, metric validation, and traffic draining. The automated CI/CD workflow must satisfy the following criteria: Route a small percentage (5%) of live user requests to a canary backend group. Run automated canary analysis for 15 minutes by evaluating Cloud Monitoring metrics and health checks against defined Service Level Objectives (SLOs). Immediately redirect traffic away from the canary and gracefully drain active requests without dropping connections if an anomaly is detected. Automatically promote the canary to 100% traffic and orchestrate graceful connection draining on the retiring legacy backend MIG before deprovisioning. Which strategy should you integrate into your CI/CD pipeline to orchestrate this automated traffic draining and canary workflow?

Configure URL map weighted backend service routing on the load balancer, use pipeline automation to adjust traffic percentages, evaluate canary health metrics via Cloud Monitoring, and leverage Backend Service Connection Draining timeouts during rollback or final cutover.: What URL Map Weighted Routing and Connection Draining Is URL map traffic splitting (weighted backend services) is a native Google Cloud Application Load Balancer feature that allows incoming HTTP(S) traffic to be distributed across multiple backend services according to configurable percentage weights. Connection draining is a backend service setting that ensures in-flight TCP/HTTP requests are given a configured duration (e.g., 300 seconds) to complete cleanly when a backend service or instance group is removed or its weight is reduced to zero. How it Meets the Requirements Canary traffic redirection: The CI/CD pipeline uses automation (via Terraform, gcloud CLI, or the Google Cloud API) to update the URL map route rules, directing 95% of traffic to the baseline backend service and 5% to the canary backend service. Automated canary analysis: During the 15-minute observation window, the pipeline queries Cloud Monitoring metrics (such as HTTP 5xx error rates, latency percentiles, and custom metrics) to verify SLO compliance. Rapid, zero-drop rollback: If metrics breach thresholds, the pipeline immediately adjusts the canary weight back to 0%. The Application Load Balancer stops routing new requests to the canary while honoring active connections through Connection Draining. Automated promotion and legacy draining: Upon successful validation, the pipeline increases canary traffic to 100%, sets legacy backend service weight to 0%, and waits for the connection draining timeout to elapse before tearing down the legacy Managed Instance Group. Key Advantages Global control-plane orchestration: Traffic shifts take effect globally within seconds without altering client-side DNS records. Zero packet/request loss: Configurable draining timeouts guarantee active transactions and WebSocket sessions finish gracefully. Native SRE integration: Easily paired with Cloud Monitoring alerting policies, health check telemetry, and automated rollback scripts in Cloud Build or Jenkins. Why This Is Optimal This approach decouples infrastructure provisioning from traffic shifting, providing precise Layer 7 traffic routing control and automated draining natively supported by Google Cloud's network data plane.

StudyCerts

Professional Cloud DevOps Engineer

Advanced

Prepare and test your skills

Track Readiness

Set Exam Goal

Advanced

Prepare and test your skills

Track Readiness

Set Exam Goal

Automate Orchestrate Traffic Drainage Canary Pipeline

Section: Draining/redirecting traffic

Your organization operates a global, mission-critical application hosted on Google Cloud using Compute Engine Managed Instance Groups (MIGs) behind an external Application Load Balancer. To improve release velocity and prevent customer-facing incidents, the SRE team requires the deployment pipeline to automate canary deployments, metric validation, and traffic draining.

The automated CI/CD workflow must satisfy the following criteria:

Route a small percentage (5%) of live user requests to a canary backend group.
Run automated canary analysis for 15 minutes by evaluating Cloud Monitoring metrics and health checks against defined Service Level Objectives (SLOs).
Immediately redirect traffic away from the canary and gracefully drain active requests without dropping connections if an anomaly is detected.
Automatically promote the canary to 100% traffic and orchestrate graceful connection draining on the retiring legacy backend MIG before deprovisioning.

Which strategy should you integrate into your CI/CD pipeline to orchestrate this automated traffic draining and canary workflow?

Configure a Managed Instance Group rolling update with proactive updates, setting maxSurge to 5% and maxUnavailable to 0%, letting the instance group manager handle automated instance replacement and health check validation.

Configure Cloud DNS weighted round-robin routing policies with a 5-second TTL pointing directly to the external IP addresses of individual Compute Engine instances, using CI/CD scripts to adjust DNS record weights and drain instances.

Configure URL map weighted backend service routing on the load balancer, use pipeline automation to adjust traffic percentages, evaluate canary health metrics via Cloud Monitoring, and leverage Backend Service Connection Draining timeouts during rollback or final cutover.

Deploy an iptables redirection sidecar script on all Compute Engine VMs, executing CI/CD remote SSH commands during deployment to locally route 5% of incoming traffic to a newly deployed canary container port.

Professional Cloud DevOps Engineer

Advanced

Prepare and test your skills

Track Readiness

Set Exam Goal

Advanced

Prepare and test your skills

Track Readiness

Set Exam Goal

Automate Orchestrate Traffic Drainage Canary Pipeline

Section: Draining/redirecting traffic

The automated CI/CD workflow must satisfy the following criteria:

Route a small percentage (5%) of live user requests to a canary backend group.
Run automated canary analysis for 15 minutes by evaluating Cloud Monitoring metrics and health checks against defined Service Level Objectives (SLOs).
Immediately redirect traffic away from the canary and gracefully drain active requests without dropping connections if an anomaly is detected.
Automatically promote the canary to 100% traffic and orchestrate graceful connection draining on the retiring legacy backend MIG before deprovisioning.

Which strategy should you integrate into your CI/CD pipeline to orchestrate this automated traffic draining and canary workflow?

Explanations

What URL Map Weighted Routing and Connection Draining Is

URL map traffic splitting (weighted backend services) is a native Google Cloud Application Load Balancer feature that allows incoming HTTP(S) traffic to be distributed across multiple backend services according to configurable percentage weights. Connection draining is a backend service setting that ensures in-flight TCP/HTTP requests are given a configured duration (e.g., 300 seconds) to complete cleanly when a backend service or instance group is removed or its weight is reduced to zero.

How it Meets the Requirements

Canary traffic redirection: The CI/CD pipeline uses automation (via Terraform, gcloud CLI, or the Google Cloud API) to update the URL map route rules, directing 95% of traffic to the baseline backend service and 5% to the canary backend service.
Automated canary analysis: During the 15-minute observation window, the pipeline queries Cloud Monitoring metrics (such as HTTP 5xx error rates, latency percentiles, and custom metrics) to verify SLO compliance.
Rapid, zero-drop rollback: If metrics breach thresholds, the pipeline immediately adjusts the canary weight back to 0%. The Application Load Balancer stops routing new requests to the canary while honoring active connections through Connection Draining.
Automated promotion and legacy draining: Upon successful validation, the pipeline increases canary traffic to 100%, sets legacy backend service weight to 0%, and waits for the connection draining timeout to elapse before tearing down the legacy Managed Instance Group.

Key Advantages

Global control-plane orchestration: Traffic shifts take effect globally within seconds without altering client-side DNS records.
Zero packet/request loss: Configurable draining timeouts guarantee active transactions and WebSocket sessions finish gracefully.
Native SRE integration: Easily paired with Cloud Monitoring alerting policies, health check telemetry, and automated rollback scripts in Cloud Build or Jenkins.

Why This Is Optimal

This approach decouples infrastructure provisioning from traffic shifting, providing precise Layer 7 traffic routing control and automated draining natively supported by Google Cloud's network data plane.

View source documentation

Professional Cloud DevOps Engineer

Automate Orchestrate Traffic Drainage Canary Pipeline

Professional Cloud DevOps Engineer

Automate Orchestrate Traffic Drainage Canary Pipeline

What URL Map Weighted Routing and Connection Draining Is

How it Meets the Requirements

Key Advantages

Why This Is Optimal

Related questions