AIAI Agent Dev Labaiagentdev-lab.com
Sign in to save progressSign in
Menu

SAP-C03 · LESSON 5

Operational Excellence and Automation20% · 3 topics

Review IaC changes and drift; promote scanned artifacts through approvals. AI revisions require quality and safety evaluations beyond HTTP success. Monitor SLOs and user experience, and bound remediation targets, attempts and duration to prevent cascading failures.

5.1Control and automate changes and operations

Concept

Review IaC changes and drift; promote scanned artifacts through approvals. AI revisions require quality and safety evaluations beyond HTTP success. Monitor SLOs and user experience, and bound remediation targets, attempts and duration to prevent cascading failures.

How it works / steps

  1. Write requirements and success criteria first.
  2. Use a test environment and retain success and failure evidence.
  3. Review the result, then check your decisions with domain practice below.

Workplace example

Test rollout stop conditions and rollback, retaining evidence for operators.

Common confusions and exam tips

Mini quiz

Q1. An ECS service is deployed blue/green with CodeDeploy. If errors increase in the new version, you want to revert to the old version without human intervention. Which method is most appropriate?

  1. Stop blue/green, switch to all-at-once deployment, and re-release a fix immediately when errors occur
  2. Greatly increase client-side retry counts so errors are absorbed by retries
  3. Have staff check logs hourly after deployment and redeploy the old task definition if errors are high
  4. Associate CloudWatch alarms with the deployment group and enable automatic rollback on alarm
Show answer and explanation

Answer: D Associate CloudWatch alarms with the deployment group and enable automatic rollback on alarm

If an alarm fires during deployment or the wait period, CodeDeploy can automatically roll back and return traffic to the original task set.

Q2. Production settings were changed manually in the console, diverged from the staging environment and caused an outage. What is the most appropriate improvement?

  1. Record config changes with AWS Config and notify staff on change
  2. Record change actions in CloudTrail to identify who made them
  3. Codify the configuration in CloudFormation, change via templates, and find manual changes with drift detection
  4. Require approval of change requests before console changes in production
Show answer and explanation

Answer: C Codify the configuration in CloudFormation, change via templates, and find manual changes with drift detection

Manage configuration as code and make changes through reviewed templates. CloudFormation drift detection finds differences between the template and the actual configuration. [Other options] Config or CloudTrail records find changes but don't keep environments matching. Approval alone still leaves manual-change mismatches.

Reference links

5.2Design a deployment strategy to meet business requirements

Concept

Deployment strategies include in-place, rolling, blue/green, canary and linear. To eliminate downtime and roll back quickly, use blue/green (Route 53, ALB weighting, CodeDeploy); to validate with limited impact, use canary. Use CloudFormation StackSets to deploy the same foundation to multiple accounts and Regions, and CodePipeline for application pipelines.

How it works / steps

  1. Confirm requirements for acceptable downtime and rollback speed
  2. Choose a deployment method that fits the requirements
  3. Define infrastructure as IaC and deploy it to all organization accounts with StackSets
  4. Switch gradually with CodeDeploy, Lambda aliases or ECS blue/green
  5. Roll back automatically when CloudWatch alarms detect anomalies

Workplace example

Release a new payment API version as a 10% canary with a Lambda alias, monitor the error rate for 10 minutes, and switch fully if there are no problems.

Common confusions and exam tips

Mini quiz

Q1. You are choosing a release method for a new version. Being able to roll back to the old version within minutes if there is a problem is the top priority, and paying for two environments during the switchover is acceptable. Which is the most suitable?

  1. All-at-once deployment
  2. Canary deployment
  3. Blue/green deployment
  4. Rolling deployment
Show answer and explanation

Answer: C Blue/green deployment

Blue/green prepares a separate new environment (green) and switches traffic. If there is a problem, you just send traffic back to the old environment (blue).

Q2. You are choosing how to codify and deploy infrastructure. You want to deploy the same IAM roles and Config rules to all accounts and Regions in an OU, and automatically apply them to accounts added in the future. Which is the most suitable?

  1. CloudFormation change sets
  2. CloudFormation StackSets
  3. AWS CDK
  4. AWS Service Catalog
Show answer and explanation

Answer: B CloudFormation StackSets

StackSets create and update stacks in multiple accounts and Regions from one template. With service-managed permissions, they also deploy automatically to accounts added to the OU.

Reference links

5.3Develop a strategy to improve operational excellence

Concept

The core of operational improvement is automation and observability. Automate routine tasks (restarts, patching, AMI creation) with Systems Manager Automation runbooks, and visualize state with CloudWatch (metrics, logs, alarms, Synthetics) and X-Ray (distributed tracing). Detect incidents with EventBridge and standardize response with Incident Manager or similar.

How it works / steps

  1. Identify manual operating procedures and automate them as runbooks
  2. Monitor key metrics with CloudWatch dashboards and alarms
  3. Identify slow services and dependencies with X-Ray
  4. Trigger automated responses from events with EventBridge rules
  5. Prevent recurrence through post-incident reviews

Workplace example

When disk utilization exceeds 90%, EventBridge triggers Systems Manager Automation to delete old logs and posts the result to Slack.

Common confusions and exam tips

Mini quiz

Q1. Which statement about AWS X-Ray is incorrect?

  1. Tracing can be enabled with integrations such as Lambda and API Gateway
  2. Sampling rules adjust the proportion of requests recorded
  3. A service map shows how requests flow across multiple services and the latency of each segment
  4. It automatically records the contents of all traffic without code changes or installing an SDK or agent
Show answer and explanation

Answer: D It automatically records the contents of all traffic without code changes or installing an SDK or agent

Using X-Ray requires SDK instrumentation, installing ADOT or the X-Ray daemon, or enabling tracing on the service side. It records trace information, not the full contents of all traffic.

Q2. An existing system with steadily increasing usage may hit service quota limits by next year. What is the most appropriate measure?

  1. Contact AWS Support to discuss after errors occur from hitting the limit
  2. Freeze the number of users at the current level and do not let new departments use the system
  3. Monitor Service Quotas utilization with CloudWatch and request increases before approaching limits
  4. Quotas increase automatically with usage, so no monitoring or preparation is needed
Show answer and explanation

Answer: C Monitor Service Quotas utilization with CloudWatch and request increases before approaching limits

Watch usage trends and raise quotas before reaching them. Some Service Quotas utilization metrics can have alarms set.

Reference links

Lessons are written by this site based on public information and do not guarantee exam content. Product specifications change, so check the latest information at the reference links.

Notes (unofficial, original questions, no pass guarantee)