[{"id":"awsadvanced-001","courseId":"aws-advanced","domain":"compute-and-events","topic":"Lambda SnapStart","prompt":"A team enables SnapStart on a Java function. Its init code generates a random ID and stores it in a static field. Many new environments are restored from one snapshot. What is the risk?","hint":"Think about when the snapshot is taken relative to your init code, and what it saves.","explain":"SnapStart saves the memory and disk state after initialization, then restores that one saved state into many environments. Anything your init code created before the snapshot is reused along with it.","source":"https://docs.aws.amazon.com/lambda/latest/dg/snapstart.html","interaction":{"type":"single"},"options":[{"id":"a","text":"Every restored environment may hold the same ID, because they all start from one snapshot.","why":"The ID was made during init and saved in the snapshot, and one snapshot seeds many environments, so the ID can repeat."},{"id":"b","text":"Lambda re-runs init for each restored environment, so every environment gets a new ID.","whyNot":"Restored environments resume from the snapshot instead of initializing from scratch, so the stored ID doesn't change."},{"id":"c","text":"Encrypting the snapshot gives each environment its own copy of the static field.","whyNot":"Encryption protects the stored snapshot; it has no effect on the values the snapshot already contains."},{"id":"d","text":"Only the first environment gets the ID, and the rest start with the static field empty.","whyNot":"Restored environments start from the same saved state, so the field is not empty in them."}],"answer":"a","explanation":"SnapStart uses one snapshot of the initialized state for many execution environments, so unique content created during init can repeat unless it is regenerated after init.","concept":"SnapStart uses one snapshot of the initialized state for many execution environments, so unique content created during init can repeat unless it is regenerated after init.","rev":1,"updatedAt":"2026-10-23","kind":"scenario","takeaways":["Generate unique IDs and secrets after init, not before the snapshot is taken."]},{"id":"awsadvanced-002","courseId":"aws-advanced","domain":"compute-and-events","topic":"Lambda provisioned concurrency","prompt":"You can configure provisioned concurrency on a published version or an alias of a Lambda function, but not on the $LATEST version.","hint":"Think about which kinds of function version can be pre-initialized.","explain":"Provisioned concurrency keeps execution environments initialized ahead of requests, so they respond right away. Before you rely on it, check which kind of function version the setting accepts.","source":"https://docs.aws.amazon.com/lambda/latest/dg/provisioned-concurrency.html","interaction":{"type":"true-false"},"options":[{"id":"a","text":"True","why":"The docs attach provisioned concurrency to a version or alias and state that it cannot be used with the $LATEST version."},{"id":"b","text":"False","whyNot":"The statement is accurate: $LATEST cannot use provisioned concurrency, while published versions and aliases can."}],"answer":"a","explanation":"Provisioned concurrency pre-initializes execution environments for a version or alias; it cannot be used with the unpublished $LATEST version.","concept":"Provisioned concurrency pre-initializes execution environments for a version or alias; it cannot be used with the unpublished $LATEST version.","rev":1,"updatedAt":"2026-10-23","kind":"scenario","takeaways":["Point the event source at the alias or version that has provisioned concurrency, or it won't use those environments."]},{"id":"awsadvanced-003","courseId":"aws-advanced","domain":"compute-and-events","topic":"Spot Instance interruptions","prompt":"Your Spot Instances in a fleet stop without warning. Which TWO of these are reasons Amazon EC2 may interrupt a Spot Instance?","hint":"Think about why EC2 would reclaim spare capacity, and which Spot request settings can trigger a stop.","explain":"Spot Instances use spare EC2 capacity, so EC2 can reclaim them when it needs that capacity back. Only a few triggers cause this, and some depend on settings in your Spot request.","source":"https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/spot-interruptions.html","interaction":{"type":"multiple","tellCount":true},"options":[{"id":"a","text":"EC2 needs the capacity back to repurpose it for other demand.","why":"Capacity is the main reason: EC2 reclaims Spot Instances mainly to repurpose capacity, and also for host maintenance or decommissioning."},{"id":"b","text":"The Spot price falls below the maximum price in your request.","whyNot":"A price under your maximum is not a listed trigger; the trigger is the Spot price rising above it."},{"id":"c","text":"The instance has run continuously for more than 24 hours.","whyNot":"Running time is not a listed trigger; interruptions come from capacity, price, or constraints."},{"id":"d","text":"The Spot price rises above the maximum price in your request.","why":"Price is a listed reason: an interruption can happen when the Spot price is higher than your maximum price."},{"id":"e","text":"Your account reaches its On-Demand vCPU quota.","whyNot":"Account quotas are not among the listed reasons; the page names capacity, price, and constraints."},{"id":"f","text":"You stop the instance yourself from the console.","whyNot":"That is a user action, not EC2 reclaiming capacity. An interruption is EC2 taking the instance back."}],"answer":"ad","explanation":"EC2 interrupts a Spot Instance when it reclaims capacity, when the Spot price exceeds your maximum price, or when a launch or Availability Zone constraint can no longer be met.","concept":"EC2 interrupts a Spot Instance when it reclaims capacity, when the Spot price exceeds your maximum price, or when a launch or Availability Zone constraint can no longer be met.","rev":1,"updatedAt":"2026-10-23","kind":"scenario","takeaways":["Setting a maximum price can make your instances be interrupted more often."]},{"id":"awsadvanced-004","courseId":"aws-advanced","domain":"compute-and-events","topic":"AWS Nitro System","prompt":"A network engineer reads that Nitro features carry forward between versions, so they assume a Nitro v5 instance keeps every feature of v4. Which statement correctly describes Nitro feature support across versions?","hint":"Check whether the carry-forward rule has any stated exceptions, and read each version's own list.","explain":"Nitro versions each list their network features, and newer versions usually keep the features of earlier ones. Read the entry for the exact version you run, since a version's own list can differ from the rule.","source":"https://docs.aws.amazon.com/ec2/latest/instancetypes/ec2-nitro-instances.html","interaction":{"type":"single"},"options":[{"id":"a","text":"Newer Nitro versions keep every feature of earlier versions, with no exceptions.","whyNot":"The page qualifies the rule: features carry forward except where a version explicitly states otherwise."},{"id":"b","text":"Newer Nitro versions keep the features of all earlier versions, unless a version explicitly states otherwise.","why":"The page says Nitro features are cumulative, so newer versions support prior features except where explicitly stated otherwise."},{"id":"c","text":"Each Nitro version lists only its own new features, so nothing carries forward between versions.","whyNot":"The page calls Nitro features cumulative, so earlier features carry forward unless a version says otherwise."},{"id":"d","text":"Only Nitro v4 and earlier carry features forward; v5 and later start a fresh feature list.","whyNot":"The cumulative rule is not limited to older versions; it covers newer versions such as v5 and v6."}],"answer":"b","explanation":"Nitro features are cumulative: newer versions support the features of all prior versions, except where a version explicitly states otherwise.","concept":"Nitro features are cumulative: newer versions support the features of all prior versions, except where a version explicitly states otherwise.","rev":1,"updatedAt":"2026-10-23","kind":"scenario","takeaways":["Feature carry-forward has explicit exceptions, so check each version's own list."]},{"id":"awsadvanced-005","courseId":"aws-advanced","domain":"compute-and-events","topic":"Standard vs Express Step Functions workflows","prompt":"A Standard workflow charges a card in a task whose state definition includes a Retry block. The first attempt fails and Step Functions retries it. Which statement about that task's execution is correct?","hint":"Think about which setting in the state definition can make a task run again.","explain":"Step Functions workflow types make different promises about how often a task runs. Settings in a state definition can cause a task to run again, so read them alongside the promise of the workflow type.","source":"https://docs.aws.amazon.com/step-functions/latest/dg/choosing-workflow-type.html","interaction":{"type":"single"},"options":[{"id":"a","text":"The task can run more than once, because its Retry block is an exception to the exactly-once model.","why":"The page says Standard tasks are never run more than once unless Retry behavior is specified in ASL, so the retry is the exception."},{"id":"b","text":"Exactly-once still holds, because Step Functions blocks the retry from repeating the charge.","whyNot":"The page gives no such blocking; its exactly-once statement carves out tasks with Retry behavior specified."},{"id":"c","text":"The task runs at most once, because retries follow the at-most-once model.","whyNot":"Standard workflows follow exactly-once, not at-most-once; at-most-once is not the Standard guarantee."},{"id":"d","text":"The at-least-once model applies, because retries are a feature of Express workflows.","whyNot":"The workflow type is fixed when you create the state machine and sets the model, so a retry does not switch it to Express."}],"answer":"a","explanation":"Standard workflows follow an exactly-once model, where tasks are never run more than once unless Retry behavior is specified in ASL.","concept":"Standard workflows follow an exactly-once model, where tasks are never run more than once unless Retry behavior is specified in ASL.","rev":1,"updatedAt":"2026-10-23","kind":"scenario","takeaways":["A Retry block in a Standard workflow can make a task run more than once."]}]