A team creates a full-size copy of its production environment on demand, runs a load test against it, and then deletes it the same day. Which Well-Architected general design principle does this illustrate?
Intermediate · Choose one
Explanation
Answer: A. The principle "test systems at production scale" says to create a production-scale test environment on demand, complete the testing, and then decommission the resources. Because you pay only while the environment is running, you can simulate the live environment for a fraction of the cost of maintaining a permanent on-premises test estate.
B. Drive architectures using data means collecting data on how architectural choices affect workload behavior and making fact-based decisions. The load test may produce such data, but the principle illustrated by creating and deleting the environment is testing at production scale.
C. Improve through game days means scheduling simulated events in production to test how the architecture, processes, and people respond. This scenario is a scheduled load test on a disposable copy, not a simulated production incident.
D. Consider evolutionary architectures means allowing the design to change over time as innovations become available, supported by automation and on-demand testing.
12
An operations team schedules a regular exercise in which it deliberately simulates a failure event in production to see how the architecture, the runbooks, and the on-call engineers respond. Which Well-Architected general design principle is this?
Intermediate · Choose one
Explanation
Answer: D. Game days are described as regularly scheduled exercises that simulate events in production so that the team can see how the architecture and its processes actually perform. They surface gaps that documentation review never finds and build organizational experience in handling real events.
A. Testing at production scale means standing up a full-size environment on demand for testing. A game day is run against the live production system and its people.
B. Automating with experimentation in mind is about creating and replicating workloads cheaply, tracking changes to automation, and reverting when needed.
C. Stop guessing your capacity needs concerns scaling capacity to demand rather than forecasting it in advance.
13
A team defines its infrastructure and its operational runbooks as version-controlled code, and triggers those procedures automatically in response to events. Which operational excellence design principle does this reflect?
Intermediate · Choose one
Explanation
Answer: B. Performing operations as code means applying the same engineering discipline used for application code to the whole environment: define the workload as code, script operational procedures, and trigger them automatically in response to events. AWS notes that this limits human error and produces consistent responses to events.
A. Refining operations procedures frequently means looking for opportunities to improve procedures as they are used and as the workload evolves.
C. Learning from all operational failures means driving improvement from lessons learned and sharing them across teams.
D. Anticipate failure means running pre-mortem exercises to identify potential sources of failure and testing the response procedures for them.
14
A company adopts a policy that all releases must be small, deployed frequently, and reversible if they cause problems. Which pillar and design principle does this policy support?
Intermediate · Choose one
Explanation
Answer: A. This is the operational excellence design principle "make frequent, small, reversible changes." Small increments make it far easier to identify which change caused a problem, and reversibility means a bad release can be withdrawn quickly rather than fixed under pressure, so the flow of beneficial changes into the workload increases without raising risk.
B. Testing recovery procedures is about validating that failure recovery works, typically by simulating failures before they happen in earnest.
C. Preparing for security events concerns incident management policy, investigation processes, and response simulations.
D. Measuring overall efficiency compares the business output of a workload against the cost of delivering it. It is a financial measure, not a release practice.
15
An architecture review recommends replacing one very large database server with several smaller instances behind a distribution layer, so that the failure of any one instance removes only part of the capacity. Which reliability design principle does this follow?
Intermediate · Choose one
Explanation
Answer: B. Scaling horizontally replaces one large resource with multiple smaller ones so that a single failure does not take down the whole workload. Distributing requests across many small units also shrinks the blast radius of any one failure, which raises aggregate availability.
A. Automatically recovering from failure means monitoring key performance indicators and triggering automated recovery when thresholds are breached. That is a response mechanism rather than the structural change described here.
C. Managing change through automation means applying changes to infrastructure through automation so that the change itself is tracked and reviewable.
D. Testing recovery procedures means validating that the recovery path works, typically by simulating failures. It does not change the topology of the workload.
16
A development team stops running its own search cluster and instead consumes a managed AWS search service, so that its engineers no longer need specialist expertise to operate it. Which performance efficiency design principle does this illustrate?
Intermediate · Choose one
Explanation
Answer: C. Democratizing advanced technologies means delegating complex tasks to the cloud vendor and consuming the technology as a service rather than asking the team to learn to host and run it. AWS gives NoSQL databases, media transcoding, and machine learning as examples of capabilities that become consumable services, which frees the team to focus on product development.
A. Mechanical sympathy means choosing the technology approach that aligns best with your goals, for example matching a data store to the workload's access patterns.
B. Experimenting more often means using virtual, automatable resources to run comparative tests across instance types, storage options, and configurations.
D. Go global in minutes means deploying into multiple AWS Regions to lower latency for customers worldwide.
17
A team selects a purpose-built database whose data model matches the workload's access patterns instead of forcing every workload onto one general-purpose engine. Which performance efficiency design principle does this reflect?
Intermediate · Choose one
Explanation
Answer: C. Mechanical sympathy is the principle of using the technology approach that aligns best with your goals. AWS gives the example of considering data access patterns when selecting a database or storage option, which is precisely the reasoning behind choosing a purpose-built engine over a default one.
A. Using serverless architectures removes the need to run and maintain servers for traditional compute activities. That is a different lever, though it may also be appropriate.
B. Go global in minutes is about deploying across Regions to reduce latency for a worldwide user base.
D. Democratizing advanced technologies is about consuming complex technology as a managed service instead of building expertise to run it.
18
A finance team and an engineering team jointly build capability in forecasting, budgeting, and cost awareness so that cloud spending is understood and controlled as a discipline. Which cost optimization design principle does this describe?
Intermediate · Choose one
Explanation
Answer: D. Implementing cloud financial management means investing in this discipline the way an organization invests in security or operations: dedicating time, people, knowledge, programs, and processes so that the business becomes cost efficient. It is about building an organizational capability rather than performing a single optimization.
A. Adopting a consumption model means paying only for what you use and increasing or decreasing usage with business need, for example stopping development environments outside working hours.
B. Stopping undifferentiated heavy lifting means letting AWS carry data center operations and managed service burdens so the team can focus on customers and business projects.
C. Analyzing and attributing expenditure means identifying the cost and usage of workloads and assigning them transparently to revenue streams and workload owners.
19
A company tags its resources so that each workload's cost can be reported to the team that owns it, letting those owners see their own spending and act on it. Which cost optimization design principle does this support?
Intermediate · Choose one
Explanation
Answer: B. Analyzing and attributing expenditure is about identifying cost and usage accurately and attributing IT costs transparently to revenue streams and individual workload owners. That visibility is what lets owners measure return on investment and act to optimize their own resources.
A. Measuring overall efficiency compares the business output of a workload against the cost of delivering it, which is a ratio rather than an attribution exercise.
C. Implementing cloud financial management is the broader organizational capability. Attribution is one practice within it, and the question describes that specific practice.
D. Adopting a consumption model concerns paying only for resources consumed and scaling usage with business requirements, such as turning off idle non-production environments.
20
A team consolidates several lightly used workloads onto fewer, better-utilized instances, noting that two hosts running at 30 percent utilization consume more energy than one host running at 60 percent. Which sustainability design principle does this apply?
Intermediate · Choose one
Explanation
Answer: A. Maximize utilization is the principle that calls for rightsizing workloads and implementing efficient designs so the underlying hardware runs at high utilization. AWS uses exactly this comparison, noting that two hosts at 30 percent are less efficient than one at 60 percent because of the baseline power each host draws regardless of load.
B. Reducing downstream impact means lowering the energy or resources customers need to use your service, including avoiding forcing them to upgrade their devices.
C. Establishing sustainability goals means setting long-term targets such as reducing the compute and storage required per transaction, and planning so that growth reduces impact intensity.
D. Understanding your impact means measuring the impact of the workload, modeling future impact, and establishing key performance indicators from that data.