Pair your devices with a code and playback position follows you: pause on this device, hit resume on the other. Position is saved to the site every minute and on pause.
Open this panel on your other device and enter the same code.
How to use this file. Each drill is a pair of near-identical answer options and a one-line requirement. Cover the right-hand explanation. Say out loud which word differs, what that word commits you to, and which option survives. Then check.
This trains lesson 2's skill in isolation: read only the difference.
⚠️ The option pairs are original, written for this course. They're not from the AWS Official
Practice Question Set. Every fact used to decide a pair is quoted from the page in sources:, fetched
2026-09-25.
D1. Requirement: offload read-heavy reporting from the primary.
Changed words: Multi-AZ / read replica. Y survives. "a Multi-AZ standby cannot serve read requests. Multi-AZ deployments are designed to provide enhanced database availability and durability, rather than read scaling benefits" (RDS FAQ). X is impossible, not just weaker.
D2. Requirement: no committed transaction may be lost if the primary's AZ fails.
Changed word: synchronously / asynchronously. X survives. Multi-AZ "uses synchronous replication between primary and standby". Read replicas use "native, asynchronous replication" (RDS FAQ). Asynchronous means a replica can lag the primary.
D3. Requirement: cross-Region DR where the secondary must take writes in under a minute.
Changed words: RDS read replica / Aurora global database. Y survives. Aurora global database can "promote one of the secondary regions to take read/write responsibilities in less than one minute". For DB instances other than Aurora, "the process takes a few minutes to complete and rebooting is part of the process" (DR options).
M1. Requirement: three consumers must each receive every message, and one is offline for hours.
Changed words: endpoint / SQS queue per consumer. Y survives. Fanout to SQS queues is AWS's named pattern ("replicated and pushed to multiple endpoints, such as … Amazon SQS queues", SNS), and the queue holds messages for a retention period configurable "from 1 minute to 14 days" (SQS FAQ). X has no buffer.
M2. Requirement: a payment must never be processed twice, and the consumer cannot be changed.
Changed word: standard / FIFO. Y survives. "Standard queues provide at-least-once delivery… FIFO queues provide exactly-once processing." With standard queues "you must design your applications to be idempotent" (SQS FAQ), and the stem forbids changing the consumer.
M3. Requirement: a workflow waits for a human approval that may take two days.
Changed word: Express / Standard. Y survives. "Express Workflows only support Request Response integrations" and "can run for up to five minutes". Standard workflows "can run for up to one year" (Step Functions). X fails twice.
M4. Requirement: orchestrate a very high event rate of short IoT processing steps, cost-sensitive.
Changed word: Standard / Express. Y survives. "Express workflows are ideal for high-event-rate workloads, such as streaming data processing and IoT data ingestion" (Step Functions). It's the mirror image of M3. The same pair flips depending on the requirement.
C1. Requirement: a job that runs about 40 minutes, no servers to manage.
Changed words: Lambda / ECS on Fargate. Y survives. Lambda: "Up to 15 minutes per invocation" (Lambda). Fargate "removes the need to choose server types" (Fargate).
C2. Requirement: register a Fargate-backed ECS service with a load balancer target group.
instance."ip."Changed word: instance / ip. Y survives. "you must choose ip as the target type, not instance. This is because tasks that use the awsvpc network mode are associated with an elastic network interface, not an Amazon EC2 instance" (Fargate).
N1. Requirement: active-passive DNS failover to a standby site.
Changed word: weighted / failover. Y survives. "You configure active-active failover using any routing policy (or combination of routing policies) other than failover, and you configure active-passive failover using the failover routing policy" (Route 53).
N2. Requirement: a partner must allowlist fixed IP addresses for your load balancer.
Changed words: fixed Elastic IP / resolved DNS name. X survives. NLB: "you can optionally associate one Elastic IP address per subnet" and "Support for static IP addresses for the load balancer" (NLB). A DNS name doesn't give the partner a fixed address to allowlist.
S1. Requirement: no user, including the root user, may delete records during the retention period.
Changed word: governance / compliance. Y survives. Compliance: "can't be overwritten or deleted by any user, including the root user in your AWS account." Governance: "users can't overwrite or delete an object version or alter its lock settings unless they have special permissions" (Object Lock).
S2. Requirement: protect against an engineer accidentally deleting data.
Changed words: replication alone / plus point-in-time backups. Y survives. "Continuous replication of data has the advantage of being the shortest time (near zero) to back up your data, but may not protect against disaster events such as data corruption or malicious attack (such as unauthorized data deletion) as well as point-in-time backups" (DR options).
R1. Requirement: the DR Region must serve requests immediately at reduced capacity.
Changed words: switched off / running. Y survives. "pilot light cannot process requests without additional action taken first, whereas warm standby can handle traffic (at reduced capacity levels) immediately" (DR options).
R2. Requirement: the recovery Region must survive a failover even if the control plane is degraded.
Changed words: rely on Auto Scaling / full capacity in advance. Y survives. "Because Auto Scaling is a control plane activity, taking a dependency on it will lower the resiliency of your overall recovery strategy. It is a trade-off. You can choose to provision sufficient capacity such that the recovery Region can handle the full production load" (DR options). ⚠️ If the stem instead says "most cost-effective" with no control-plane requirement, X can be the answer. The requirement decides.
Q1. You see "Amazon EMR" in an option and don't recognise the expansion.
Y. "Not every abbreviation is fully spelled out on the exam or available in the Help feature. The official full name for some AWS services includes an abbreviation that is never expanded (for example, Amazon API Gateway, Amazon EMR)" (service mentions).
The fastest way to lock in a pair is to write it. For every fact you learn in SAA1–SAA4:
If you can't find the quote in step 4, you don't know the fact well enough yet. That's worth knowing before the exam, not during it.