Study guide
13.2 Rollout Strategies: A/B Tests, Canaries & Traffic Splits
Google Cloud Professional Machine Learning Engineer PMLE
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: Exam Overview & the Agent Platform Map
1.1 PMLE Exam Facts, Policies & the June 2026 Blueprint 1.2 From Vertex AI to Gemini Enterprise Agent Platform: The Product Name Map
2 Chapter 2: Low-Code ML with BigQuery ML
2.1 Choosing, Training & Predicting with BigQuery ML Models 2.2 Feature Engineering & Feature Selection in BigQuery ML 2.3 Fine-Tuning & Evaluating Gemini Models from BigQuery
3 Chapter 3: AutoML & Google Cloud AI APIs
3.1 Training Models with Agent Platform AutoML 3.2 Building with Document AI, Vision, Translation & Other AI APIs
4 Chapter 4: Foundation Models in Model Garden
4.1 Evaluating & Selecting Models in Model Garden 4.2 Building & Tuning Solutions with Gemini, Imagen, Veo & MaaS 4.3 Optimizing Gemini Applications for Cost, Latency & Availability
5 Chapter 5: Exploring, Preprocessing & Protecting Data
5.1 Organizing & Exploring Tabular, Text & Image Data 5.2 Choosing a Preprocessing Tool: BigQuery, Dataflow, Spark or Python 5.3 Creating & Consolidating Features in Agent Platform Feature Store 5.4 Protecting PII & Sensitive Data in ML Workflows
6 Chapter 6: Notebooks, Prototyping & Experiment Tracking
6.1 Workbench vs. Colab Enterprise: Secure, Collaborative Notebooks 6.2 Prototyping with PyTorch, scikit-learn, JAX & Model Garden in Notebooks 6.3 Experiment Environments, Tracking & Lineage with Experiments and ML Metadata
7 Chapter 7: Evaluating Predictive & Generative AI Solutions
7.1 Evaluation Metrics for Predictive Models 7.2 Evaluating Gen AI: Computation Metrics, Rubrics & LLM-as-a-Judge
8 Chapter 8: Choosing the Modeling Approach
8.1 Choosing the Model Type: ARIMA, Trees, DNNs & LLMs 8.2 Choosing the Product & Deployment Strategy 8.3 Modeling Techniques for Interpretability Requirements
9 Chapter 9: Training Data & Training SDKs
9.1 Organizing & Ingesting Training Data from Cloud Storage & BigQuery 9.2 Custom Training on Agent Platform: Containers, Jobs & Worker Pools 9.3 Kubeflow on GKE, AutoML & Tabular Workflows as Training Choices
10 Chapter 10: Troubleshooting, Hyperparameter Tuning & Fine-Tuning
10.1 Troubleshooting ML Training Failures 10.2 Hyperparameter Tuning on Agent Platform 10.3 Fine-Tuning Foundation Models & When to Tune
11 Chapter 11: Training Hardware & Distributed Training
11.1 Choosing CPUs, GPUs & TPUs for Training 11.2 Distributed Training: Data & Model Parallelism
12 Chapter 12: Deploying Models for Inference
12.1 Batch vs. Online Inference on Agent Platform 12.2 Serving Beyond Agent Platform Endpoints: Model Garden, Cloud Run & GKE 12.3 Prebuilt & Custom Serving Containers
13 Chapter 13: Model Versions, Rollouts & Inference Logic
13.1 Organizing & Versioning Models in Model Registry 13.2 Rollout Strategies: A/B Tests, Canaries & Traffic Splits 13.3 Inference Preprocessing & Postprocessing
14 Chapter 14: Scaling Online Serving
14.1 Serving Features Online with Feature Store 14.2 Public, Dedicated & Private Endpoints 14.3 Serving Hardware & Autoscaling for Throughput 14.4 Tuning Models for Production Training & Serving
15 Chapter 15: End-to-End ML Pipelines
15.1 Validating Data & Models in Pipelines 15.2 Agent Platform Pipelines with KFP: Components, Templates & Caching 15.3 Ensuring Consistent Preprocessing Between Training & Serving
16 Chapter 16: Orchestration with Managed Airflow & Ray
16.1 Orchestrating ML Workflows with Managed Service for Apache Airflow 16.2 Ray on Agent Platform for Distributed ML Workloads
17 Chapter 17: Automating Retraining & CI/CD/CT
17.1 Choosing a Retraining Policy 17.2 CI/CD/CT for ML with Cloud Build
18 Chapter 18: Securing AI Systems, Responsible AI & Explainability
18.1 Protecting AI Systems from Data & Model Exfiltration 18.2 Defending Gen AI Apps: Safety Filters, Model Armor & Regex Guardrails 18.3 Responsible AI: Detecting & Monitoring Bias 18.4 Model Explainability on Agent Platform
19 Chapter 19: Monitoring AI Solutions
19.1 Configuring Model Monitoring on Agent Platform 19.2 Training-Serving Skew, Data Drift, Concept Drift & Attribution Drift 19.3 Monitoring, Testing & Evaluating Gen AI Solutions in Production
13.2 Rollout Strategies: A/B Tests, Canaries & Traffic Splits
Key Takeaways Explain This Topic AI
An Agent Platform endpoint can host several deployed models and split traffic between them by percentage, and the percentages must total 100.
A canary rollout sends a small share of traffic, such as 5-10%, to a new version and increases it only while monitored metrics stay healthy.
An A/B test splits comparable traffic between versions for long enough to measure a business KPI with statistical confidence.
A rolling deployment (preview) replaces a deployed model with a new version on the same compute and reverts if new replicas never pass health checks.
Rollback on an endpoint means shifting the traffic split back to the previous deployed model, which stays deployed during the rollout.
Last updated: September 2026
The exam guide lists implementing model rollout strategies (for example, A/B testing and canary deployments) to compare model versions . On Agent Platform, the main mechanism is the traffic split on an endpoint.
Traffic Splitting on Endpoints
An endpoint can have several deployed models (different models or versions) at once. The trafficSplit map assigns a percentage of requests to each deployed model ID, and the percentages must add up to 100 . When deploying a new model, the key "0" refers to the model being deployed in that request: "trafficSplit": {
- "0": 10,
- "DEPLOYED_MODEL_ID_CURRENT": 90
- }
Each deployed model has its own replicas and machine configuration , so the new version gets its own capacity. Traffic splits work on dedicated public, shared public, and Private Service Connect endpoints. Private endpoints that use VPC Network Peering don't support traffic splitting.
Rollout Strategies Compared
Strategy How it works on Google Cloud Goal Rollback
Canary Deploy the new version to the same endpoint with a small split (for example, 5%), watch metrics, then step up (25% → 50% → 100%) Catch problems with minimal user impact Set the canary's split to 0%
A/B test A steady split (for example, 50/50) between champion and challenger for a planned duration. Log which version served each request and join with outcomes Measure business impact (conversion, fraud loss) with statistical confidence End the test and route 100% to the winner
Blue/green Deploy the full new version (green) alongside the current one (blue), validate, then switch 100% at once Fast, clean cutover with instant fallback Switch traffic back to blue
Rolling deployment (preview) Replace a deployed model with a new version on the same compute , moving traffic gradually as replicas become ready Update without extra capacity Automatic revert if new replicas never become healthy
Shadow evaluation Send copies of production inputs to the new model offline (for example, replay logged requests through batch inference) without serving its outputs Compare predictions safely before any user impact Nothing to roll back
Rolling deployment details
A rolling deployment (set with rolloutOptions ) creates a new DeployedModel with a new ID and revision number . It reuses the previous deployment's traffic split and dedicated resources, and it replaces replicas as surge capacity allows. Traffic moves in proportion to ready replicas. If new replicas never pass health checks , no traffic goes to them, and the deployment fails and reverts to the previous model.
Designing a Canary
Define guardrail metrics before starting: error rate, p95 latency, prediction distribution (for example, share flagged as fraud), and business proxies.
Start small (1-10%) with enough traffic to see problems quickly.
Compare canary and control on the same time window, because traffic patterns change by hour and day.
Step up only when the guardrails hold for a set period.
Automate rollback : alert policies on endpoint metrics trigger a pipeline or script that sets the canary split to 0%.
- Designing an A/B Test
- Hypothesis and KPI: "The new ranking model increases add-to-cart rate by at least 2%."
Randomization unit: a split per request can put one user in both arms. For user-facing experiences, assign users consistently in the application layer (for example, by hashing the user ID) and call the matching deployed model, or use separate endpoints.
Sample size and duration: run long enough to reach significance and cover weekly cycles.
Logging: enable request-response logging, record the model version with each prediction, and join with outcome data in BigQuery.
Delayed labels: fraud or churn outcomes may arrive weeks later, so plan the analysis window.
What to Monitor During Rollout
- Layer Metrics
- Serving health Request count, error rate (4xx/5xx), latency percentiles, replica count, CPU/GPU utilization
- Model behavior Prediction distribution, confidence scores, input feature distributions (skew and drift)
- Business Conversion, approval rate, manual review volume, complaints
- Cost of Safe Rollouts
Each deployed model on an endpoint has its own replicas , so a canary or blue/green rollout temporarily pays for two deployments. Keep this cost down by sizing the canary's minimum replicas for its traffic share, setting a clear end date for A/B tests, and undeploying the old version once rollback is no longer needed. A rolling deployment avoids double capacity by reusing the same compute, at the cost of less side-by-side comparison.
Gen AI Rollouts
The same patterns apply to prompt or model changes in gen AI applications:
Canary a new prompt template or model version to a small share of users.
Run Gen AI evals on sampled production traffic for both variants (Chapter 7).
Watch safety filter block rates , cost per request, and latency, along with quality.
Worked Scenario
A bank deploys fraud model v8 next to v7 on a Private Service Connect endpoint.
Deploy v8 with trafficSplit {"0": 5, "v7": 95} and minimum replicas sized for 5% of peak.
For 48 hours, compare decline rate, manual-review volume, p95 latency, and error rate between the versions.
Raise to 25%, then 50%, then 100% as guardrails hold.
Keep v7 deployed at 0% for one week as an instant rollback path, then undeploy it to save cost.
Move the champion alias in Model Registry to v8.